Image compression system and method
FIELD: physics, computer engineering.
SUBSTANCE: invention relates to image compression systems and methods. The method of compressing a digital image in a computing device comprises steps of dividing an image into a plurality of image subregions; selecting from a catalogue which includes a plurality of predetermined template forms, wherein each template form comprises a plurality of elements, properties and image variables, such as colour, colour gradient, gradient direction or reference pixel, and wherein each said form is identified by a code, the template form of each subregion best corresponding to one or more image elements of said subregion; and generating a compressed data set for the image, wherein each subregion is represented by a code which identifies the template form selected therefor.
EFFECT: improved compression of image data, thereby reducing the amount of data used to display an image.
22 cl, 4 dwg
The technical field to which the invention relates
The present invention relates to systems and methods for image compression, in particular to reduce the amount of data used to represent images.
The level of technology
The formation of digital images can create digital images, typically from a physical object. The digital image may be created directly from a physical scene by a camera or similar devices. Alternatively, the digital image may be obtained from the other image on analog media such as photographs, photographic film, or printed document, and can be converted into a form of digital media via a scanner or similar device. Many technical images, such as obtained with the aid of such topographic equipment, devices for scanning computed tomography (CT), radar side view or radio telescopes, are obtained through a complex process, not related to the image data. Finally, the digital image can also be calculated using a geometric model or mathematical formula.
The digital image may include pixels. The pixel may be the smallest component information in from�the image. Typically, the pixels are arranged in a two-dimensional lattice with a regular structure and are often represented using dots, squares or rectangles. Each pixel can have a value that represents a property of the sample source image. Generally, increasing the density of pixels or samples provides a higher resolution or more accurate representation of the original image. The intensity of each pixel can be variable. In TV color systems, each pixel can have three or four color components such as red, green and blue or cyan, Magenta, yellow and black.
Image resolution can evaluate the quality of the image. Image resolution can be described by means of the size of a pixel in the digital image. The image has a size in integer number N of pixels in height by an integer M pixels wide can have any resolution less than or equal to NxM pixels (covering N lines per picture height, or N lines TV). One other common rule describes the resolution as the total number of pixels in the image, usually specified as a certain number of megapixels, which can be calculated by multiplying the columns of pixels in rows of pixels (NxM) and dividing by one million. Other rules describe� resolution the number of pixels per length unit or pixels per area unit, such as pixels per inch or per square inch. These computed pixels resolution, in General, are inaccurate (the true resolution can be less than the calculated permissions, and can serve as confidants or upper boundary values of the true resolution of the image. In General, the higher the resolution, the more detail appears in the image.
The pixels may be stored in a storage device of the computer in the form of a raster image or bitmap textures or bit-array: a two dimensional array of small integers. Often, these values are transmitted or stored in compressed form. Each pixel of the bitmap as a rule, is associated with a certain "position" in a two-dimensional (2D) area of the image and the values of one or more image elements in this position. Digital images can be classified according to the number and properties of these samples pixels.
Image compression can reduce redundancy of the image data to reduce the amount of image information that will be stored or transmitted. Compression of the image may be "lossy" (if the recovered data after compression will differ from the original, resulting in data loss) or "lossless" (if the recovered data after compression is the same as that Isho�governmental data). These losses can be used, if the lost data is sufficiently small or if the benefit of reducing the data outweighs the damage caused by data loss. The lossless data compression allows to recover the exact original data from the compressed data. Lossless compression can be used, if identity is an important source and recovered data compression or if the importance of an exact copy or permissible deviation from them is not known. The usual types of data, compressed by lossless data compression, are executable programs and source code. For some file formats, such as portable network graphics (PNG) or graphics interchange format (GIF), usually is used only lossless compression, while others, such as the bitmap format (TIFF) graphics and network with multiple images (MNG), can be used as lossy and lossless.
The compression mechanisms may require different performance for encoding and decoding. The compression method is often measured by the peak signal-to-noise. The peak signal-to-noise can measure the number of errors or noise introduced in the process of image compression with losses. However, subjective judgment is maturing�La is also seen as an important and perhaps the most important to evaluate the accuracy of compression.
Disclosure of the invention
There is an unmet need for a system and method of image compression, which largely correspond to the information content of the image itself and also provide a large compression of the image data, and their presence would be beneficial.
Embodiments of the invention can overcome the shortcomings of the prior art by dividing the image into multiple subareas and each subarea of the image through one or more forms similar to a master from among a directory or database is already existing template forms the image.
Brief description of the drawings
The subject invention which is regarded as the invention is specifically characterized and definitely claimed in the concluding portion of the specification. However, the invention both the organization and method of operation, together with its objectives, characteristics and advantages, may be best understood by reference to the detailed description when considered with the accompanying drawings. Specific embodiments of the present invention will be described with reference to the following drawings, in which:
Fig.1 is a schematic drawing�her system for image compression in accordance with a variant implementation of the present invention;
Fig.2 is a schematic illustration of data structures for image compression in accordance with a variant implementation of the present invention;
Fig.3 is a block diagram of a method of compression of image data for one or more image frames in accordance with a variant implementation of the present invention; and
Fig.4 is a block diagram of a method of recovery after compression of the image data to restore one or more source image frames in accordance with a variant implementation of the present invention.
It will be understood that, for simplicity and clarity of the image, the elements depicted in the drawings are not necessarily shown to scale. For example, the dimensions of some elements may be enlarged relative to other elements for clarity. In addition, if appropriate, a reference position can be repeated in the drawings to indicate corresponding or analogous elements.
The implementation of inventions
The present invention consists of a system and method reduce the size of images using compression.
Embodiments of the invention can divide the image into many sub-regions and represent each sub-region of the image to one or more of such boilerplate forms, chosen from a catalogue or database�nnyh existing template forms image. For example, if your image contains primarily shades of green, such as in the scene with the forest containing trees and grass, to represent parts of the image can be used green template. In some cases, the entire image can be marked shared (e.g., average) of template form, despite the fact that the individual pixels of the image data are represented by referring to this template form (or, in some cases, a variety of template forms). In such embodiments, instead of each pixel in the form of an absolute criterion (e.g., red, blue or green) of each pixel value can evaluate the property (for example, level "red," "blue" and/or "Telenesti") regarding (e.g., red, blue and/or green) template forms. Because the form template is approximately equal to the original image, many pixels so can exactly match a template form that no image data could not be used to represent these pixels. According to such variants of implementation, to represent the image can be used a smaller amount of data compared to the original image data, and therefore their transmission and preservation can be more effective.
Such variants of implementation can od�way be applied to a whole image or part of an image, where many groups of pixels described by a constant or a defining attribute functions, such as, for example, a template form with color. The image can be divided into subareas in a number of ways. For example, the image matrix can be linearly separated for presentation to a smaller group of pixels within the image, for example, where each sub-matrix image (3 x 3 can represent the sub-region. Known algorithms for dividing images, for example, among other things, "k-means (k-means) or other algorithms for selecting areas or sub-areas of pixels, in some cases, can be used to identify sub-areas within the image.
The image can be divided into many sub-areas, each sub-region forms a submatrix of the matrix of the image and, therefore, represents a subregion of pixels. For example, each sub-matrix can represent a square, rectangle or other polygon, consisting of many pixels. As a non-limiting example, each square can, in some cases, include values for subarea size 10x10 pixels.
Once the image is divided into many subdomains, the subdomains can be checked with the color or other image elements. Proven e�COP image could be the most dominant element in the subarea. For example, if the image element is associated with the shape with a particular color, it can be checked the most dominant color in this subarea. May be the comparison of the dominant color (or other element) with a lot of these colors (or other image elements) stored in the directory or in the database of picture elements. The data processing unit can automatically determine which of the colors in the catalog (or other image elements) most closely match the colors (or elements) for each subarea. Alternatively, for each subarea can be checked a variety of colors, if the process can be performed in multiple colors and their relative(s) location(s).
Then from the directory can be selected form template with a representation of the image element that most closely matches the performance for each subarea. Then all pixels in that sub-region can be replaced with the corresponding pixel values of this pattern forms. A template form can be represented by the number in the directory (e.g., code, address, or other reference marker that uniquely identifies a master form or function in the catalog). In some embodiments, to represent or encode the plurality of pixels�fishing in each subarea can be sent stored or used in any other way only one number in the directory, thereby resulting in a substantial decrease in the amount of data required to represent a pixel in each sub-region of the image.
In some cases, can be used in any form or number of pixels for each subarea; however, the smaller the size of each subarea, the greater the number of subdomains used in the image. Such an increase in the total number of subdomains and decreasing the size of each subarea can increase the resolution in pixels and to ensure greater accuracy relative to the original image, but also can increase the amount of data required to represent the original image. The parameters defining the size, shape and/or density of the subareas of the image can be configured by the user or, alternatively, may be automatically configured or installed by the computer processor (for example, to support pre-defined maximum allowed data to represent the image and/or a pre-defined minimum correctness or accuracy allowed for the submission of the original image).
In accordance with some embodiments of implementing the present invention, each template form may include�ü in values for many elements, properties and variables of the image. The elements or variables of the image may include, among other things, color, color gradient, the direction of the gradient or of the reference pixel.
In some embodiments, the compressed data representing the pixels can simply be a reference to one or more of such boilerplate forms or functions of the truth values from the catalog. The compressed data can be "lossy", if a template of the form do not match with the original image exactly. In another embodiment, the implementation, the formation of compression lossless each pixel may be assigned an additional value of the difference (or "error") indicating the difference between the template form(s) and the source image. Thus the original image can be exactly reconstructed (decoded) through the use of forms and then add the values of errors for each pixel.
Reverse operation for decoding compressed data in which the image is reconstructed from a template form that can be executed by a processor or decoder. The decoder can extract the template form or the closest representation of the attribute functions of the image from the catalog or database (e.g., identified by the code in the directory). The decoder may use the form(s) (and, in some �instances, data errors) to restore the original image. If you use lossy compression, any difference between the form template and the original image data is lost through compression. If you use lossless compression, the difference or error values can be recorded for each pixel and sent to the decoder, for example, with the data template or separately from them to restore the original image without loss of data relative to the difference between the form template and the source image.
When using lossy compression the closer match of the template form and the original image, the less data is lost. If you use lossless compression, the closer the match the template form and the original image, the less data errors are stored and transmitted with the compressed image data (a perfect match leads to the absence of data errors). Accordingly, to improve the accuracy lossy compression or data reduction errors are used in the lossless compression, for the best match with the original image during the compression process, the directory may be adjusted. In one embodiment, the implementation may be provided with a set of modified directories with additional or modified template form(s). For example, for the transmission of Fig�message catalog or only the changed parts can be sent periodically, at least more than once as the compression and sending of the image.
For example, suppose that the initial image is represented by a matrix, whose dimensions are 9x9 pixels, which gives a General picture of the 81 pixel. This image may be divided, for example, a group of nine subregions, and the hypermatrix of 3x3 pixels. Each sub-region can be characterized by the appropriate form, which includes one or more of the mentioned elements or image variables. The closest template shape can be selected from the catalog by comparison with many of these boilerplate forms. Alternatively, if no form in the catalog does not show significant benefits compared to another, it can be used a template the default form, e.g., having the smallest amount of data, such as a monochrome pattern with an average numeric value of the color. In another variant implementation, if the directory is not present sufficiently close pattern forms, then the directory can be added to the template form. New form template can replace the previously existing form in the catalog or can be added as additional forms. To replace existing forms already existing form that needs to be replaced can be selected in accordance�accordance with one or more criteria, which can include, at least, those that were not previously used (or were used the least often) to represent the subarea of the image. Additionally or alternatively, a form that needs to be replaced, can be the shape that is closest to the new form or to the form that is most different from other already existing forms in the catalog, for example, to provide a wider range of options for the template.
In some embodiments, the processor may automatically determine which of the template forms in the directory is "closest" to the original image or sub image. "Proximity" boilerplate forms to the original image or sub image may be determined in accordance with any type known in the prior art form or function of the comparison images. In some embodiments, a greater selection can be set based on proximity of some elements (e.g., color of individual pixels) unlike other elements (for example, the overall color gradient and/or intensity, or Vice versa). For example, the processor may classify or evaluate each template using a weight function comparison in which to compare various elements can be given different weights Il� priorities.
As soon as the characteristic form template(s) selected to represent each subarea of the image, the sub-region can be represented by a number, code or address in the directory corresponding to this template(am). In some cases, the compressed data can include the size or dimensions of the subarea of the image; however, if the image is divided into equal size sub-region, requires the inclusion in the composition of this descriptor sizes or can be included in their composition with the data once, or every time you change the dimensions of the subdomains. In addition, the compressed data may include relative location of each sub-region within the original image; however, if a subarea of the image compressed in accordance with a predefined or expected order, then their membership should not be included the location of each specific sub-areas.
If the compressed data is described by a set of image elements, each element can be represented as coordinates in the vector. The vector may include not only the number (index) of color, but also other information such as, for example, the starting color, gradient color, gradient color and a reference label of the pixel. Consequently, the initial image 81 of the pixel is divided into nine podobna�TEI of 3x3 pixels, can be described by 9 "vectors pixels", with each vector defines the values of the image for a set of selected items for each subarea. Combining the different values of the elements in the vectors may reduce the total amount of storage capacity and processing load on the data-processing units, for example, reducing the number of variables treated by approximately 45%, from 81 (9 by 9) 36 (9 x 4). The data size can be further reduced, if sequential images are processed, for example, in a short period of time. Each subsequent image may be a function of the input variables of basis vectors; hence, over time, as each image reduces the number of variables treated by approximately 45%, this reduction can be increased for each processed image. To further reduce the size of the image data of each sub-region of the image can be represented by a simple binary form using only two colors (for example, instead of three or four colors).
In accordance with some embodiments of the invention, instead of determining the above-mentioned set of one or more variables for each subarea separately process can also ensure�th many forms, and then to choose the form that is appropriate for each sub-region or group of subfields. The form provided in the catalogue may be pre-determined or fixed and/or formed or at least adjusted during image compression.
The number of template forms, is available for image compression, may be proportional, of the same order or relatively high compared to the number of individual pixels, the number of rows of pixels or any size of pixels in the image. In one example, the number of template forms may be approximately equal to the square root of the number of pixels in the sub-region (e.g., subareas 256x256 pixel can have 256 available forms). You may also have other or additional number of forms. Although the increase in the number of specimens, as a rule, increases the size of the directory, and the increase may also increase the accuracy of the images compressed using the forms.
Each form may include information of images, such as color gradients and/or other elements of the image reference pixel, the direction of the gradient, etc. For each sub-region from catalog forms can be chosen form. For each sub-region can also be selected reference color, for example, may be the average color pixel� in subareas. Forms can be built before processing or image compression to ensure the previously described directory that, for example, may be at least partially determined heuristically. If the directory is not detected shape sufficiently similar to the image data for sub-regions, it can be applied standard technology "subsampling" to reduce the size of data subareas. In another embodiment of the implementation, if no sufficiently similar shapes, data for subareas may remain uncompressed.
Alternative or additionally, can be formed a new form in order to sufficiently match the subregion of the image, and it can be added to the forms directory. For example, the new form may be exact or modified copy of the subareas. The data processing unit may set a threshold value to determine which forms are "sufficiently similar", for example, in accordance with comparison of the parameters for the form. The higher the similarity threshold, the more accurate can be compression of the image. In some embodiments, the viewer can describe or detail the standard precision image compression by comparing the restored images of compressions with time�ary thresholds of similarity with the original image and the selection, for example, by visual observation, which of restored images and, hence, the similarity thresholds are acceptable.
The local data processing unit may divide and compress the images described in this document method, and then may transmit and/or store compressed image data. Then the local data processing unit can restore the image to get the full image to display.
Fig.1 schematically illustrates a system 100 for image compression in accordance with a variant implementation of the present invention.
The system 100 may include one or more devices 150, which may be removed and/or head image, the base data 110 to save the template directory forms and/or compressed images, the module 120 is split into subdomains to split each image into many subareas, the module 130 compression to compress data, each sub-region and the module 140 recovery after compression to restore the image from the compressed data subareas.
The device 150 may be a computing device, a server device for capture or playback of video or images, mobile device or any other digital device such as a cellular phone, palmtop computer�computer (PDA), console for playing video games, etc., the Device 150 may include any device capable of performing a series of commands to perform the recording, preservation, storage, processing, editing, display, projection, reception, transmission or use or manage in any other way the image data or video. The device 150 may include a device 155 output (e.g. monitor, projector, a screen, a printer or a display device) for displaying video data or image in the user interface. The device 150 may include a processor 157. The processor 157 may include a Central processing unit (CPU), digital signal processor (DSP), a microprocessor, a controller, a chip, a microchip, field programmable gate arrays (FPGA), a specialized integrated circuit (ASIC) or other integrated circuit (IC), or any other suitable multi-purpose or specialized processor or controller.
It should be understood that each base 110 of the data, the partitioning module 120, the module 130 of compression and/or module 140 recovery after compression can be an integral part of the device 150 (e.g., part of the same single computer) or be separate from the device part 150 (attached by means of a wired or wireless network).
The module 120 R�of zmienia into subdomains can break the image into many subareas, where each sub-region can be described by means of one or several defining elements, such as color, and, in some cases, one or more corresponding variable values for subareas. The module 120 is split into subdomains can share the picture in many ways. For example, the image matrix representing the values of the pixels of an image can be linearly separated into smaller groups of pixels within the image, for example, on many of the hypermatrix, each of which represents a subregion of the image. Can be used known mechanisms for dividing images, for example, among other things, "k-means (k-means)" or other methods for clustering of pixels, to identify sub-areas within the image.
Once subarea of the image described, the module 130 compression may choose a characteristic shape for each subarea, for example, from the catalog of template forms stored in the database 110 data. Then the module 130 compression can compress subareas. Each form template may include information about the image such as color gradient, the reference pixel, the direction of the gradient, etc. For each sub-region can be selected as the reference color, which may be, for example, average, medium or mild color pixel�in the subarea.
Compressed image data can be stored in the database 110. If the device 150 requests the restored image, the module 140 recovery after compression can restore a compressed image data from the database 110 data and can restore data after compression to the exact or approximate copy of the original image.
The module 120 is split into subdomains can share images, the compression module 130 can compress the image, and the module 140 recovery after compression can restore the image using the local data processing unit, such as a processor 157. Then the local data processing unit may transmit the restored image to the controller output to display it on the device 155 output.
Fig.2 schematically illustrates the form data to the compressed image in accordance with a variant implementation of the present invention. Form data can include the directory 202 subdomains with a lot of template 208 forms to represent multiple sub-areas 210 in the source image 200. Pattern forms 208 may be pre-defined and formed, for example, heuristically before compressing the image. Alternative and/or additionally, a template of the form 208 can be generated dynamically, for example, real-time�Yeni, for each image or multiple images using the current image and/or one or more preceding images. The compression module (e.g., module 130 compression from Fig.1) can use the function 204 classification and compression to select a template form 208 from the catalog 202 to compress each sub-region of the original image 200. The compression module may use the module 204 classification and compression to select one or more template forms 208 from the catalog 202, which most closely match the image elements, including color or other characteristic attributes of the function from one or more of the plurality of sub-areas 210 of the source image 200. The compression module can represent each sub-region 210 by selecting a template form (s) 208 to form compressed image 206. Compressed image 206 can represent each pixel (or set of pixels) in the sub-region 210 of the source image 200 by reference to the form in the directory 202 and/or a description of a color, such as the value or color code.
The computer or server (e.g., the device 150 of Fig.1) can form a template form 208 in the directory 202, which provides the range and details options template described, for example, in accordance with any of a variety of parameters, including, among about�it pre-defined or selected quantitative value of desirable precision compression, the maximum size of the compressed data for each image 206, the sub-region, square pixels, rectangle or other shapes of pixels and/or any other criteria. In some embodiments, the computer can generate a template of the form 208 in the directory 202 dynamic way, for example, using the forms of pictures are of the actual compressed images. For video stream or a sequence of such images template form 208 in the directory 202 may be adjusted to compress the current source image 200, in accordance with the image elements from previous image frames in the stream. The computer can generate a template of the form 208 in the directory 202, in accordance with both dynamic (e.g. adjustable during processing) and heuristic (i.e. pre-defined) properties.
If the directory 202 is changed, then the entire directory 202 or only the changed portions may periodically or repeatedly transmitted to the compression module, for example, from a database (e.g., database 110 data from Fig.1), which is stored in the directory 202, or directly from the computer that generates the changes. If the changed parts catalog 202 is periodically transmitted, the entire ka�log 202 may be transferred only once, for example, for each compression module, project, session, or set of images, thereby reducing the amount of data that must be transmitted.
Overall, the quality of the compression process image (for example, the degree to which the compression is lossy) can be described by means of directory 202 from which to retrieve the compression information. For example, if the compression module detects in the directory 202 stereotyped pattern 208, which coincides exactly with a specific attribute functions for each sub-region 210, the compression can be performed without a loss, while if not detected no exact match, the compression can be lossy. The degree to which compression is lossy (e.g., the amount of data loss or error image compression), may depend on how closely the template of the form 208 coincide with the original image subareas 210, and therefore, the detail and the number of template forms 208, available in the directory 202. Although the degree to which the compression is lossy or lossless, is not directly related to the size of the directory 202, as the number and detail of template forms 208 in the directory 202 increases, the size of the directory 202 is increased, and, more than likely that can be detected exact or closer match.
Fig.3 pok�Zana is a block diagram of a method of compression of image data for one or more image frames in accordance with an embodiment of the invention.
In operation 301, the database (e.g., database 110 of Fig.1) can provide a compression module (e.g., module 130 compression from Fig.1) with the initial directory having a plurality of template forms. The initial directory can be adopted, for example, even before the analysis of the first image or sub image. The computer or server (e.g., the device 150 of Fig.1) can form a template form, for example, in accordance with one or more heuristics and/or pre-defined parameters.
In operation 302, the partitioning module (for example, the partitioning module 120 of Fig.1) may receive the first image and may divide the image into multiple subfields. Subareas can be defined geometrically (for example, the first image is divided in such a way as to form a predetermined number of subareas of pre-defined sizes) or content-based image (for example, the first image is divided along the boundaries identified by significant changes of image elements, such as color).
In operation 303, the compression module may receive the first image and may perform analysis of the image elements of the image together or separately for each subarea. The compression module may use the function to�assification and compression (for example, function 204 classification and compression of Fig.2) for determining and/or selecting a template form from the home directory with the most similar image elements. If available, the most similar template form and/or sufficiently close to the first image or sub-region, to compress the image or sub-regions may be selected form template. Otherwise, the computer or the server may create a new template form that sufficiently matches the image or sub-region. A new template can be added to the initial directory in the form of additional images or can replace one or more patterns within the initial directory for forming the second directory.
In operation 304, once the master form, the compression module may compress the first image using template forms, for example, through the representation of the image or each subarea by means of the identifier of the selected form in the catalog. To form exact copies of the original image or subarea of the image (i.e., the compression is "lossless") may be transferred to the value of the difference or "error" that indicates the difference between the template form(s) and the pixels of the original image or subarea of the image, along with the ID of the selected template forms�. This difference or "error" can take many possible forms, including, for example, a two-dimensional matrix in which the value of each element represents the difference between the values of corresponding pixels in the source image or sub-regions, and selected a template form, a two-dimensional array of pixels. In this case, the sum of the two-dimensional matrix of differences and a two-dimensional array of pixels of the template may be a two-dimensional matrix that represents the set of pixels of the original image or the original subarea of the image.
In operation 305, the compression module may transmit the identifier to select a template form in the catalog, and, in appropriate cases, the difference or "error" on the module recovery after compression (for example, the module 140 recovery after compression of Fig.1) for recovery after compression of the first image frame, as described in more detail below with reference to Fig.4.
Operations 302-305 may be repeated for each additional image frame. In addition, if the directory is updated with new template form, a new form template and/or the entire updated catalog can be transferred at the end of the above process, after all the images or sub-regions coincided with template forms, or alternatively more than once during the treatment, for example, to accommodate one or more catalog updates that may occur during processing. In some embodiments, if the analysis of the total image, the process may skip the operation 302.
Fig.4 shows a block diagram of a method of recovery after compression of the image data to restore one or more source image frames in accordance with an embodiment of the invention. The method can recover the data after compression that has been compressed, for example, in accordance with the variants of implementation described with reference to Fig.3.
In operation 401, the module recovery after compression (for example, the module 140 recovery after compression of Fig.1) can extract the compressed data for the frame image. The image can be subdivided into a number of subdomains (for example, as described in operation 302 with reference to Fig.3). Module recovery after compression can perform operations 402 and 403 for each subarea of the image frame as follows.
In operation 402, the module recovery after compression can be extracted from the compressed data, the ID of the template shape, and, in appropriate cases, the values of the difference or "error" for the current subarea. The ID of the form template may include address, ssy�ku or code for a directory or location in the database storing a unique template form.
In operation 403, the module recovery after compression can use the ID to access the directory and extract the pixel values of the template shape for each pixel (or group of pixels) in the current subregion. Module recovery after compression may use the pixel values of the template form to restore subareas in accordance with these values.
To produce accurate copies of the original subarea of the image (i.e. compression lossless) module recovery after compression can also use the values of the difference or "error" that define the difference between a template form(s) and the pixels of the original image or subarea of the image, as discussed above.
In operation 404, the module recovery after compression can consistently unite all recovered after compression subareas for the formation of approximation (using only the template form) or an exact copy (using the template form, and of difference values) of the original image. Module recovery after compression may transfer the image formed on the computer or server (e.g., the device 150 of Fig.1).
In operation 405, the output device (e.g., device 155 conclusion from Fig.1) computer or server may display�to press the formed image.
If there are multiple images, such as multiple frames in the video stream, the operation may be repeated for all 403-405 for each image. Operations for all 403-405 can be operated in series or in parallel for multiple image frames.
Embodiments of the invention may include a product, such as machine-readable or processor, a storage medium or a storage medium of a computer or processor, such as, for example, a storage device, a floppy drive, a USB (universal serial bus) flash memory, for storing commands that, when run by a processor or controller (e.g., processor 157 of Fig.1), carry out methods disclosed herein.
Although the invention has been described regarding a limited number of embodiments, it is clear that can be made many changes, modifications and other applications of the invention.
1. The method for compressing a digital image in a computing device, the method contains the stages at which:
divide the image into many sub-areas of the image;
choose the directory that includes many pre-defined template forms, and each form template contains a set of elements, properties, and variables of the image, such as color, zvetovospriatie, the direction of the gradient or of the reference pixel, and each of the mentioned form template is identified by a code, a template form for each sub-region that most closely matches one or more image elements in the subregion; and
form a compressed data set for the image, in which each sub-region is represented by the code identifying the selected template for him the form.
2. A method according to claim 1, in which the directory is defined before first analyzed sub-region.
3. A method according to claim 1, wherein the directory is adjusted whenever the analyzed one or more subdomains.
4. A method according to claim 1, wherein, if none of the plurality of template forms are not in the directory corresponds to a subarea of the image sufficiently, then a new form template and added to the directory to compress this subarea of the image.
5. A method according to claim 1, wherein the image is divided geometrically to form a predetermined number of subareas of pre-defined sizes.
6. A method according to claim 1, wherein the image is divided based on the contents of the image.
7. A method according to claim 6, in which the boundaries between the subdomains coincide with the boundaries of the image content that contains significant changes in the elements of the image�.
8. A method according to claim 1, comprising stages on which to choose a variety of template forms for each subarea and each form is chosen for the different image element.
9. A method according to claim 1, wherein the compressed image can be restored after compression using code to retrieve a form template from the catalog for each subarea, restore the full image by combining the template forms for all of subfields and displays the restored image.
10. A method according to claim 1, wherein each sub-region in addition is represented by the value of the difference representing the difference in one or more image elements between the subregion of the image and selected for this template form, so that form data lossless compression.
11. A method according to claim 10, wherein the compressed image can be restored after compression using code to retrieve a form template from the catalog for each subarea and the values of differences for one or more elements of an image, restore an exact copy lossless full image by combining the template forms and values of difference for all of subfields and displays the restored image.
12. The compression system that contains:
a storage device for storing rolled�and, contains many pre-defined template forms, with the mentioned pattern forms contain many elements, properties, and variables of the image, such as color, color gradient, the direction of the gradient or of the reference pixel, and each mentioned template form is identified by the code; and
a processor for dividing the image into multiple subfields, selecting a template form from the catalog that most closely corresponds to the image element for each sub-region and forming a compressed data set for the image, in which each sub-region of the image represented by the code identifying the selected template for him the form.
13. A system according to claim 12, further comprising a calculating device for determining the directory to the analysis of the first subarea of the image processor.
14. A system according to claim 12, further comprising the computing device to adjust the directory, when analysing one or more of processor subdomains.
15. A system according to claim 12, in which, if the processor determines that none of the plurality of template forms does not coincide with the subregion of the image sufficiently, the processor generates a new template form and the storage device adds a new form template to the directory to compress the p�oblasti image.
16. A system according to claim 12, in which the processor is geometrically divides the image into a predetermined number of subareas of pre-defined sizes.
17. A system according to claim 12, in which the processor divides the image based on the image content.
18. A system according to claim 17, in which the processor identifies the boundaries of the image content that contains significant changes in the elements of the image, and automatically divides the image along the boundaries of the image content.
19. A system according to claim 12, in which the processor selects the set of template forms for each subarea so that each form is most closely matched for different subareas of the image element.
20. A system according to claim 12, containing:
the processor to restore the image, and the processor uses the code from the compressed data set to extract a template of the form in the directory for each sub-region and restores the full image by combining the template forms for all of the subareas; and
a display device for displaying the restored image.
21. A system according to claim 12, in which the processor is additionally each sub-region by the difference in one or more image elements between the sub-region and selected for this template form, so that FD�data is approximated by a lossless compression.
22. A system according to claim 21, comprising:
the processor to restore the image, and the processor uses the code from the compressed data set to extract a template form from the catalog and the values of differences for one or more elements of the image for each subarea and restores the full image without any loss by combining of template forms and of difference values for all of the subareas; and
a display device for displaying the reconstructed image without loss.
FIELD: physics, computer engineering.
SUBSTANCE: invention relates to computer engineering. A data processing apparatus comprises an information data input unit; an information data encoding unit; a unit for generating a file with information, having a plurality of block storage elements, each including information on size, and which comprises one or more first block storage elements for storing encoded information data, and one or more second block storage elements for storing control information on information data stored in the first block storage element; a unit for setting a mode in accordance with an instruction from a user of one of a plurality of modes, including a recording mode for recording the file with information on a recording medium and a transmission mode for transmitting the file with information to an external device, wherein the method includes, in transmission mode, generating a first block storage element each time the amount of encoded information data reaches a predetermined amount, generating a second block storage element corresponding to the generated first block storage element and transmitting the first block storage element after transmitting the second block storage element.
EFFECT: facilitating simultaneous reproduction of received information data.
4 cl, 10 dwg
FIELD: information technologies.
SUBSTANCE: in this method they accept a largest coding unit (LCU) of video data divided into a set of coded units (CU) with a small size of the unit in accordance with the scheme of quadrant tree division; they determine whether the CU includes non-zero coefficients of transformation by means of decoding of coded data for reproduction of indication of the fact whether the CU includes non-zero coefficients of transformation; syntax elements are decoded for the CU, to indicate the change in the quantisation parameter for the CU relative to the predicted parameter of quantisation, if CU includes non-zero coefficients of transformation, besides, one or more elements of syntax are decoded from the position within limits of coded video data.
EFFECT: increased quality of data compression.
30 cl, 8 dwg
FIELD: information technology.
SUBSTANCE: image decoding device receives the feedforward encoded bit stream, which is generated by dividing each signal frame of the moving image on the support blocks of a given size. The device decodes the bit stream for receiving the moving image signal. The device comprises a module of decoding for decoding the bit stream for receiving the information indicating the given size, and the information indicating the threshold block size. The device also operates in the mode of the motion prediction and determines the motion vector for each of the support blocks, or for each of the single blocks of the motion prediction, identified as blocks received by hierarchical division of the support blocks. The motion prediction mode defines the procedure of the motion prediction for the single blocks of the motion prediction.
EFFECT: increased efficiency of decoding the information.
FIELD: physics, computer engineering.
SUBSTANCE: invention relates to rendering content with a rendering apparatus, e.g. a television set connected to DLNA (Digital Living Network Alliance). Disclosed is a rendering apparatus (130) connected to a network, which receives a command to play a content item and/or the content item to be played from another network connected device (110). Instead of automatically interrupting the current rendering activities of the rendering apparatus for playing the content item, the content item is added to a list of scheduled content items (140).
EFFECT: reducing interruptions of content being rendered on network connected rendering apparatus in case of "pushing" files thereto from other network connected devices.
13 cl, 6 dwg
FIELD: physics, computer engineering.
SUBSTANCE: invention relates to computer engineering. An image decoding device comprises an internal prediction module for, when the encoding mode for an encoding unit is an internal encoding mode, performing intra-frame prediction over each unit which is a unit for the prediction process for forming a prediction image, wherein said unit is said encoding unit or a unit obtained by dividing said encoding unit, wherein said internal prediction module generates an intermediate predicted value from reference samples according to an internal prediction parameter indicating the type of internal prediction, sets a value which is obtained by filtering the intermediate predicted value as the final predicted value only for at least one specific position in said unit, and sets the intermediate predicted value as the final predicted value for any other positions in said unit.
EFFECT: high image quality by reducing prediction errors.
4 cl, 10 dwg
FIELD: physics, computer engineering.
SUBSTANCE: group of inventions relates to predictive coding means. The method comprises receiving a quantised converted coefficient of a current unit, obtaining the internal prediction direction of the current unit, finding the scanning order corresponding to the internal prediction direction of the current unit from a table of correspondence of the internal prediction direction and the scanning order. In the method, the scanning order in the table of correspondence of the internal prediction direction and the scanning order corresponds to two internal prediction directions, the two internal prediction directions have geometric correlation along the reference internal prediction direction, geometric correlation between two internal prediction directions in said directions is symmetric correlation along the reference internal prediction direction.
EFFECT: reduced scanning orders of quantised converted coefficients of a current unit and complexity of a codec system in encoding and decoding.
12 cl, 5 dwg, 5 tbl
SUBSTANCE: method comprises, at the sender, dividing an electronic image into macroblocks, each macroblock divided into N≥2 blocks from which K<N synchronisation blocks are selected; calculating a transmission synchronisation sequence and embedding into synchronisation blocks which allow embedding; at the recipient, establishing synchronisation of the digital watermark in the received electronic image for which, beginning with the selected initial point of the recipient, the received electronic image is successively divided into macroblocks and blocks, from which blocks for the intended synchronisation are selected; extracting therefrom verification subsequences and merging into a verification sequence which is compared bitwise with all shifts of the calculated receiving synchronisation sequence; the electronic image received by the recipient is considered an electronic image with established synchronisation of the digital watermark, corresponding to the intended synchronisation sequence with the least number of mismatches. The disclosed method can be used to increase the probability of establishing synchronisation of a digital watermark of an electronic image, divided into component parts of an arbitrary size, and for preventing visually noticeable distortions caused by embedding into blocks of the electronic image with virtually unchanged statistical characteristics of transmission synchronisation sequences.
EFFECT: improved establishment of synchronisation of a digital watermark of an electronic image when dividing the electronic image with an embedded digital watermark into component parts of an arbitrary size.
4 cl, 9 dwg
SUBSTANCE: method comprises obtaining an intraframe prediction mode of a current intraframe encoding unit from a predefined set of prediction modes; obtaining reference prediction modes of the current intraframe encoding unit, wherein the reference prediction modes are intraframe prediction modes of available adjacent units relative to the current intraframe encoding unit or prediction modes in a preset reserved set of reference modes; recording a first flag bit into a code stream according to the reference prediction modes and the intraframe prediction mode; obtaining prediction mode encoding values according to the size relationship between the intraframe prediction mode value and reference prediction mode values and encoding the prediction mode encoding values. Use of the encoding method disclosed herein can provide saving on determination logic at the encoding side, thereby improving encoding efficiency.
EFFECT: high encoding and decoding efficiency.
28 cl, 5 dwg, 1 tbl
FIELD: physics, computer engineering.
SUBSTANCE: invention relates to computer engineering. A decoding method which comprises accessing a video image which includes multiple images merged into a single image, wherein the video image is part of a received video stream; accessing information which is part of the received video stream, wherein the accessed information indicates how the multiple images are merged in the accessed video image, and includes information on sampling and information on spatial interleaving, which includes information on a relationship which indicates the type of relationship existing between the multiple images, and also indicates that the multiple images are stereoscopic projections of an image, the multiple images are not linked, the multiple images are two-dimensional images and their associated depth map (2D+Z), the multiple images are multiple sets 2D+Z (MVD), the multiple images are layered depth video (LDV) format images or the multiple images are images two sets LDV (DES); and decoding the video image to provide a decoded representation of at least one of the multiple images.
EFFECT: high efficiency of decoding.
10 cl, 41 dwg, 10 tbl
FIELD: physics, computer engineering.
SUBSTANCE: invention relates to computer engineering. A method of predicting video signals comprises: obtaining a block of pixels, wherein the block of pixels includes integer pixel values corresponding to integer pixel positions within the block of pixels; calculating a first sub-pixel value for a first sub-pixel position, wherein calculation of the first sub-pixel value comprises: applying a first interpolation filter defining a first one-dimensional array of filter coefficients corresponding to filter support positions; calculating a second sub-pixel value for a second sub-pixel position, wherein calculation of the second sub-pixel value comprises applying a second interpolation filter defining a second one-dimensional array of filter coefficients corresponding to horizontal filter support positions and applying a third interpolation filter defining a third one-dimensional array of filter coefficients corresponding to vertical filter support positions, wherein the first one-dimensional array comprises more filter coefficients than the second one-dimensional array; the first one-dimensional array comprises more filter coefficients than the third one-dimensional array; and generating a prediction block based on at least the first sub-pixel value and the second sub-pixel value.
EFFECT: high prediction accuracy.
37 cl, 15 dwg
FIELD: information technology.
SUBSTANCE: image coding apparatus has a converter with increase in pixel depth in bits for converting the depth in bits of each pixel of the input image so as to output a converted input image and output information on conversion in bits indicating the number of bits for which depth changes as a result of the conversion, an image coder for encoding the converted input image so as to output image coding information and a multiplexer for multiplexing information on conversion in bits and image coding information.
EFFECT: high coding efficiency owing to high accuracy of intra-frame prediction or compensation for movement.
34 cl, 60 dwg
FIELD: information technology.
SUBSTANCE: invention can be used to transmit and receive television (TV) signals for broadcast and industrial-institutional television, generated, for example, using a TV camera or other signal sources of black and white, colour, spectrozonal, three-dimensional or some other images via interlaced or line-by-line scanning of TV images. At the transmitting side after analogue-to-digital conversion of the brightness signal E'Y and two colour difference signals E'R-Y and E'B-Y, and generating binary symbols in a k-bit parallel code, a decryption operation is performed for binary symbols of the parallel code of one reading for the brightness E'Y and two colour difference signals E'R-Y and E'B-Y and coincidence signals are generated in form of logic unit for each code combination of the k-bit parallel code of parallel symbols which characterise one signal level from m=2k for the current moment in time Δt for one reading, after which each coincidence signal in form of logic unit for the brightness E'Y and two colour difference signals E'R-Y and E'B-Y is modulated at its carrier frequency via quadrature modulation. Multiplexing is then performed for digital modulated signals of separate levels for the brightness E'Y and two colour difference signals E'R-Y and E'B-Y, as well as digital synchrosignals, and at the receiving side after demultiplexing signals, digital modulated signals of separate levels for the brightness E'Y and two colour difference signals E'R-Y and E'B-Y are picked up and then synchronously detected, after which the obtained signals are used to generate binary symbols of the k-bit parallel code of initial digital brightness signals E'Y and two colour difference signals E'R-Y and E'B-Y for each reading. Said signals then undergo digital-to-analogue conversion, after which the obtained analogue brightness signals E'Y and two colour difference signals E'R-Y and E'B-Y are amplified and then used to display the obtained video information.
EFFECT: generation of digital television signals and transmission thereof in a smaller communication channel frequency band.
2 cl, 2 dwg
FIELD: information technology.
SUBSTANCE: invention can be used to transmit and receive television (TV) signals for broadcast and industrial-institutional television, generated, for example, using a TV camera or other signal sources of black and white, colour, spectrozonal, three-dimensional or images. At the transmitting side, several analogue synchronous TV signals of gray-scale images are processed to form a brightness E'Y and two colour difference signals E'R-Y and E'B-Y in digital form through analogue-to-digital conversion thereof with sampling frequency of the brightness signal equal to a value fd, and for colour difference signals equal to a value fd/2 for multiplexing the digital brightness signal E'Y and two digital colour difference signals E'R-Y and E'B-Y presented in a serial code and separately formed digital synchronisation signals, said signals are transmitted over a communication channel, and at the receiving side inverse operations are performed over the signals compared to the transmitting side, and specifically, signals are demultiplexed, digital signals of the serial code are converted to a parallel k-bit binary code, signals undergo digital-to-analogue conversion and an analogue brightness signal EY is formed, as well as two colour difference signals E'R-Y and E'B-Y which are then displayed, wherein at the transmitting side the analogue-to-digital conversion operation over signals is performed with sampling frequency equal to fd/2 for the brightness signal and fd/4 for the colour difference signals, and the receiving side of the system, after demultiplexing digital signals, the brightness and colour difference signals are stored and read to form digital signals of a TV image with the full number of decomposition elements, wherein it is also possible that, when using interleaved scanning o the TV image at the transmitting side the analogue-to-digital conversion operation is performed with sampling frequency equal to fd/2 for the brightness signal and fd/4 for the colour difference signals without shift of clock pulses of the sampling rate in time for (1+2·m) half-frames of the odd field of the TV image and (2+2·m) half-frames of the even field of the TV image, where m=0, 2, 4, 6,…,24 and with shift of clock pulses of the sampling rate in time for duration of one decomposition element for (3+2·n) half-frames of the odd field of the TV image and (4+2·n) half-frames of the even field of the TV image, where n=0, 2, 4, 6,…,22, and at the receiving side of the system, after demultiplexing digital signals with picking up of the brightness and colour difference signals, the signals are stored and then read to form digital images of the initial odd and even fields of the TV image.
EFFECT: generation and transmission of digital television signals in a smaller communication channel frequency band.
3 cl, 2 dwg, 3 tbl
FIELD: information technologies.
SUBSTANCE: invention is related to method of video image processing, which provides for minimisation of data volume produced after compression of video image and intended for storage in accumulator or for further transfer through communication network. Video processor (1) is suggested, which is intended for comparison (7) of a separate frame contained in received video signal (3), with previous processed frame to detect variances. In order to produce compressed video information in the most compact format, processor (1) performs the following operations: detects (7) variances by distribution of separate frame pixels into movable units and identifies that movable unit as variable, in which there is a certain number of pixels, variation of colour value of which compared to appropriate pixels of previous processed frame exceeds the specified threshold, and also replaces (8) colour values of pixels in invariable movable units in processed frame for specified values and generates signal that indicates these variances.
EFFECT: minimised volume of video data for transfer or storage.
13 cl, 5 dwg
SUBSTANCE: this method comprises memorising the input raster video image as a flow of frames in the line input buffer. Said frames are splitted to micro blocs. The latter are compressed and stored in external memory. For processing, said micro blocs are retrieved from external memory, unclasped and written to internal memory. Raster macro blocs are formed and processed by appropriate processors.
EFFECT: efficient use of internal memory irrespective of processing algorithm type.
22 cl, 2 dwg
FIELD: physics, photography.
SUBSTANCE: invention relates to an image processing device and method, which can improve encoding efficiency, thereby preventing increase in load. The technical result is achieved due to that a selection scheme 71 from a prediction scheme 64 by filtering selects a motion compensation image for generating a prediction image at a high-resolution extension level from key frames at a low-resolution base level. The filter scheme 72 of the prediction scheme 64 by filtering performs filtration, which includes high-frequency conversion and which uses analysis in the time direction of a plurality of motion compensation images at the base level, selected by the selection scheme 71, in order to generate a prediction image at the extension level.
EFFECT: reducing load in terms of the amount of processing owing to spatial increase in sampling frequency at the base level for encoding the current frame.
19 cl, 26 dwg
FIELD: information technology.
SUBSTANCE: method of compression of graphic file by fractal method using ring classification of segments, in which the graphic file is split into rank regions and domains, and for each rank region the domain and the corresponding affine transformation is found, that best approximates it to the appropriate rank region, and using the obtained values of the domain parameters, comprising their coordinates, the coefficients of the affine transformations, the values of brightness and contrast, the archive is formed, and classification of domains and rank regions are introduced, based on the allocation in them of the "rings" and the calculation of the mathematical expectation of pixel intensity of these "rings", which enables to reduce the complexity of the phase of correlation of the segments and to accelerate compression.
EFFECT: reduced time of compression of the graphic file by fractal method.
SUBSTANCE: method comprises making each array element in an image sensor from one "R, G, B radiation colour brightness to code" converter, which performs parallel synchronous conversion of radiation of three colours analogue video signals R, G, B into three codes. The frame image digitisation apparatus includes an objective lens, an image sensor comprising an array of elements, three switch units, three register units and a control signal generator, wherein each switch unit includes the same number of encoders as converters.
EFFECT: reduced cross dimensions of array elements in an image sensor, which enables to reduce the frame format size or increase resolution of the image sensor.
6 dwg, 1 tbl