Method for sorting digital images for quality printing

FIELD: processing of digital images.

SUBSTANCE: method includes following operations: recording of digital images from, for example, a digital camera, into memory block, which may be a part of a computer or a printer; analysis of digital images, contained in memory block, on basis of properties: image brightness, image blur level, red eye detection; recording of metadata into files of digital images, corresponding to character of found defects; digital images, on basis of analysis results, are classified as belonging to following categories: images which satisfy high quality standards and which are usable for printing, in other words, without defects, images, usable for printing after being edited, in other words, with correctable defects, images, not usable for printing, in other words, with uncorrectable defects; in memory block, separate sections are created for storage of each one of three aforementioned categories of digital images; files of sorted digital images are recorded in memory block together with metadata about the character of found defects.

EFFECT: conservation of processing time and consumables.

6 dwg

The invention relates to the field of measurement technology, and more specifically to methods of automation selection of digital images, suitable for quality printing.

In the selection process digital photos for printing the main criterion, which uses the average consumer is the quality of the images. Some images may be unsuitable because of poor quality, i.e. to be underexposed, overexposed or have low contrast or have problematic content.

Real defects digital photos include, in particular, defects, exposure, blur and glare from the flash, noise, JPEG artifacts, color imbalance, chromatic aberration, and other defects.

As shows the analysis of digital photographic images, exposure problems are the most common defect. Follow different ways to blur and glare from the flash. On average, every tenth picture is corrupted by noise, JPEG artifacts or color imbalance. Chromatic aberration appears quite rarely. At the same time their appearance and the allocation of predictable, so to combat them can use different algorithms are adaptive search and correction. Other types of defects cannot be detected without additional special measurements and involves the motions. Often the defects are distributed in the image non-uniformly, and local correction, for example, in the case of stray light or complex color imbalance.

Currently there are several methods for automatic evaluation of image quality, allowing to determine whether the image quality is acceptable or requires improvement. In General, these methods relate to the quality of the exposition. Determining compliance of exposure, the nature of light registered object or scene is to assign the image of the following classes: underexposed, overexposed, low contrast or normal, and the choice of technique exposure correction. In the simplest case, such a technique can be a histogram stretch, positive or negative gamma correction or the prohibition of correction.

In tiled application U.S. No. 2003/0151674 [1] proposed a method for evaluating the quality of digital camera images, which includes checking the received image to determine the acceptability of the image quality and providing feedback to the user, who decides to save the image or re-photographing. The photo quality is evaluated on the basis of three characteristics: a) the sharpness of the image; b) image quality of a person is in) the presence or absence of flash in the process of shooting. Quality criteria are calculated for each of the three quality characteristics and compared with a predefined threshold to determine whether the achieved level threshold value. If all thresholds are exceeded, the user is informed that the obtained image has acceptable quality.

In tiled application U.S. No. 2005/0270381 [2] presents a way to increase the possibility of obtaining images implemented in a digital camera. The process is based on the analysis of the parameters obtained images, metadata, and comparing them with valid parameters for determining the defectiveness of the image. If no defects, there is a record of the image. The user may be informed about the problems with the image. If a glitch is detected, the opportunity to reshoot the image.

In tiled application U.S. No. 2004/0190789 [3] describes a method of characterization of exposure data based on the analysis of the global histogram of the brightness of the image, implying, in turn, the subsequent correction. Automatic exposure analysis and improvement process is divided into two main phases: the analysis phase and classification phase of the exposure correction. During the processing of the brightness histogram of the classifier selects the parameters used for analyzing the exposure image: eff the active width of the histogram, what is used to determine the range of the histogram equalization, the center and the centroid of the histogram that is used to determine whether to increase the brightness to correct underexposed images or less overexposed. Thus, correction of errors of the exposure is to a) histogram equalization; (b) increase or decrease the brightness of the image.

The autofocus digital camera does not always accurately. Pictures taken in autofocus mode, may not be in focus and, as a result, are blurred. For some photos of all the image may be out of focus due to camera movement during shooting or failure, the auto focus function, i.e. when the digital camera is unable to focus in any part of the image. In other cases, focused only part of the image. For example, the digital camera focuses on the background instead of the foreground, if the image is correctly focused. In this case, being unfocused images is a complex task, as it requires the definition of foreground and background.

Blur is one of the most important factors in the perception of image quality due to the accompanying loss of detail and sharpness of the image. Image blurring may occur su is bedstvie a large number of reasons, including, for example, defocusing lens on the object, defocusing the foreground or background due to the large values of the diaphragm, the movement of an object relative to the camera and even atmospheric haze. Blur traditional analog photographs could not be significantly reduced. However, in digital photography, image blurring can be largely eliminated. Existing techniques to reduce blurring of images typically include the evaluation unit of the blur and the unit eliminate blur. If the evaluation unit blur finds the blur of the image is above a certain threshold, the image is processed in block reduce blur. Some existing techniques reduce blur quantitatively assess the degree of blur and accordingly choose the degree of correction. Other techniques to reduce blur an image using an iterative procedure that loops over the image in the block reduce blur up until the estimated amount of blur will not be below a certain threshold.

In the publication by Doron Shaked and Ingeborg Tastl, "Sharpness Measure: Towards Automatic Image Enhancement", Hewlett-Packard Laboratories Technical Report HPL 2004-84, May 2004 [4] described an algorithm for the measurement of total clarity that defines how you want to change the image sharpness is to achieve the optimum. The algorithm evaluates the global sharpness of the image, specified by a single value for the entire image. Thus, sharpen blurry images will be increased, while quite sharp image will not be processed. The measure uses the local frequency analysis focused on the extraction of features. In this regard, the method is free from the disadvantages of alternative methods. Methods of analysis the frequency domain provide excellent measures of sharpness for images of similar scenes, however, they are unsatisfactory when scenes change. Methods focused on the extraction of features, focus on the characteristics of the images, however, the assumptions required for good performance, too strict for most common applications. The measure of sharpness correlates well with the visually perceived sharpness and largely invariant to image content. The measure of sharpness can be used to run the algorithm improvements that will increase the sharpness of the original image to the nominal value. Last, but not least, the measure of sharpness is easily feasible and requires less computation than the convolution calculation 3×3.

In U.S. patent No. 6771308 [5] describes a digital camera having a detector blur. Kam the RA itself can print photos. The camera has a built-detector, for detecting camera shake during shooting, the recording media photos and intensity of camera shake, the device for selecting the recorded pictures to be printed, the controller for analysis of images, determining whether to be printing the selected picture or not. In the future, the camera displays an alert when it thinks that the stamp of the snapshot could not be performed.

In U.S. patent No. 6298145 [6] described an image processing system including the detector blur, which determines the value of the blur image based on the information contained in the compressed image, provided that it contains a face. The detector blur informs that the image is suitable for printing and/or visual representation, if the blur value is less than a predefined threshold.

The red-eye defect continues to be the most common problem faced by users of digital cameras and imaging. It occurs due to the reflection in the eye of the object and usually manifests itself in the digital image as a reddish area where usually the black pupil of the object. Unnatural reddish gleam in the eye is due to internal reflections from the vascular membrane behind the retina, which is rich in blood vessels. This is the playing technique is called in particular, the small angle between the flash of the camera and its lens. This angle decreases with miniaturization of a camera having a built-in flash. This trend will continue with further miniaturization of cameras and the introduction of lamps flash in camera phones.

In U.S. patent No. 6252976 [7] described a computer program for detecting color defects of the eye in the image due to the flash lamp, perform the following steps: detection in a digital image regions having skin color; search areas with skin color for groups of pixels with characteristics corresponding to the defect of the red eye and adjust color pixels based on the position of the red eye.

In U.S. patent No. 6873743 [8] described an automatic detection and removal of red-eye in digital images in real time, including the detection module red eye, which is red eyes, if they are without any user intervention. If the image is found defective, the part of the image surrounding the defective area is processed by a correction module that reduces the red component of red eye, keeping the other color characteristics problematic areas. The invention minimizes the computational cost required for the detection and correction of red-eye, and thus, under the auditing for applications requiring processing in real-time large volumes of digital images to archiving or printing. This system can process the image, based on personal computers, commercial printers or digital cameras, as part of the process of obtaining, or to visualization on handheld computers, mobile phones and other digital devices.

Along with brightness and blur, color processing, color reproduction is a key factor in determining the quality of a digital image. However, understanding and modeling the visual system is far from complete, and there are several numerical models that can predict the results of color perception according to the very complex nature of our preferences. As a result of different camera manufacturers develop their own color models and approaches to get the most spectacular photos. Even for the same physical scene cameras from different manufacturers can give quite a different color on the photos because of different CCDs, processing algorithms, and used color models.

Thus, obtaining quality assessment of digital images, prepared for printing, is performed automatically on the basis of the analysis of the digital brightness of the images, the degree of blur and detection of red-eye.

After the assessment of the quality of digital images produced in a special way to register this information.

In the image file obtained by a digital camera are usually saved metadata. In the General case, metadata refers to any data associated with the file and is different from normally available to the user data of this file. In most cases, modern digital cameras use as a metadata format the format of the Exif (Exchangeable image file), the use of which is recommended by the Japan industrial Association of electronics and information technology (see Exchangeable image file format for digital still cameras: Exif Version 2.2. JEITA CP-3451 Technical Report. April 2002 [9]). Exif metadata can be considered as an additional source of information in the process of automatic processing of digital images, and contents of this information is determined by the intrinsic properties of a digital camera, and include it installed and measured parameters. The user may use a variety of available on the market software utility to view and modify metadata for the purpose of their further use, for example to improve and printing.

In addition, the automatic algorithm is able to detect low is high-quality pictures through image analysis, will give the opportunity to make the necessary information about the implemented or planned correction of the image in the metadata.

As an example, metadata is used for printing should be mentioned PDF. Its use reduces the cost and eliminates many of the problems associated with sending it to the printer for print jobs. In this case, the printer creates PDF in accordance with its specifications, passing objects, images, and fonts associated with the file.

Also known extensible metadata platform Adobe (XMP), which allows to insert metadata about the file directly in the file itself (see publications Robin L. Dale, Günter Waibel. "Capturing Technical Metadata for Digital Still Images". RLG DigiNews. Volume 8, Number 5, 2004 [10] "Adobe XMP JDF PDF: The Alphabet of Automation." Print Media. May/June 2004 [11]). For example, the digital camera automatically records information about the pictures in the image file, such as date and time, camera type, exposure time and lens. This information is metadata. Additional metadata can be added manually, such as location of the picture, the photographer's name, background information, etc. that will allow you to search for important information in connection with the shooting. Using XMP, the metadata attached to the file pictures and accompany him during the entire process. The parameters of all changes to the file, the recordings which are stated in the metadata. Even if the original file is converted to another format, such as EPS, PDF or Photoshop file, the metadata is stored, providing the file information header during the entire process.

Currently, many printers have the ability to print directly from a digital camera or memory card, but truly Autonomous of them can be named only a few. The autonomy of the printing device is determined, in our opinion, the presence of a hard disk, a set of slots for memory cards, PictBridge port, allowing to print or transfer the photo to the internal hard drive with any digital camera, as well as automatic or interactive procedure correction of digital images. It is assumed that the user can manually mark is required for printing photos, scanning them on the display. However, in the standard version, which is primarily determined by marketing (price) considerations of the manufacturer, the size of this display is small enough to accurately assess the quality of the picture. A large number of printer models and does not display, and the process of selecting images for printing takes a long time when you consider that modern memory cards digital cameras save several hundred photos at a time.

The closest is to the claimed invention is a solution, described in the posted application U.S. No. 2002/0109854 [12], in which in detail describes the method of assessing the quality of a digital image is considered as the basis for image correction and determine the number of required procedures correction, the degree of a particular procedure correction, its relative weight among other procedures, if any, and any combination of these. Each made to improve the rated maximum value which is equal to the unit. Thus, if the image is improved, i.e. edited, for one class, for example, exposure compensation, then it is assigned a rating unit; when two transformations - two rating, etc. Respectively assigned rating may depend on the type of correction and/or the degree of improvement. If the image is undergoing significant adjustments or corrections are several treatments of various types, the rating increases, for example, to two. Therefore, each image gets a rating in accordance with the degree of correction or improvement that was applied to this image. Greater improvement corresponds to a higher rating. The image is exposed in accordance with the received rating sufficient correction, admits of usable is m for printing. The highest priority for printing receives the image with the highest rating. This application shall be considered in the future as the prototype of the claimed invention.

The problem to which the invention is directed, is to develop a simple and reliable way to automatically split the available digital images on high-quality, i.e. suitable to be printed without editing, conditionally quality, i.e. suitable for printing after editing or other processing, and unfit to print the image.

The subject of the invention is an automatic way of sorting digital images, comprising the following operations:

- write digital images, for example, from a digital camera, a memory block, which may be part of the computer or printer;

- analyze digital images contained in the memory block, features:

- the brightness of the image

- the degree of blur

- detection red-eye;

- write in digital image files metadata corresponding to the nature of the detected defects;

- classify digital images, the results of the analysis into the following categories:

images that meet the highest quality standards and is suitable for printing, i.e. without the Def is tov,

images, suitable for printing, provided they are editing, i.e. with the adjustable defects

images unsuitable for printing, i.e. with fatal defects;

- create a block of memory sections for storing each of the three categories of digital images;

- write in the memory block files are sorted digital images along with metadata about the nature of the detected defects.

One of the advantages of the claimed invention is the automatic selection of the best images from a set of digital photos downloaded to the printer from its own block of memory or from the memory block Manager computer printer. Quality pictures recorded in the appropriate directory of the memory block and can be printed in batch mode without additional analysis. This gives an advantage in processing time, and conserve supplies, i.e. paper and cartridges for printing.

Features identified during the initial analysis of digital images, mark a special flag in the Exif metadata or special technical metadata, so that the images can be printed without re-classification. In addition, information about defects can be recorded in the header of the Exif (or metadata) as the defect is expozitii", "defect blur" or "red-eye defect", which will be used in the further improvement of defective images for printing.

In the process of downloading digital images from a digital camera to a storage device or in the process of analyzing existing images, for example, on the internal HDD of the printer, the classification allows you to take high quality pictures in a separate subdirectory, where the printing can be done directly or with an additional correction of the printer. The proposed method reduces the cost of consumables and time to print.

The most significant improvement that provides the claimed invention is:

- method for sorting digital images for printing with the division into three categories:

images that meet high quality standards and

suitable for printing;

- images that are not suitable for printing, but which can be

improved and further printed;

images unsuitable for printing, for which improvement

impossible.

This method reduces the consumption of paper for printing time and printing costs, increasing the efficiency of the printing process.

- Image analysis and evaluation are performed in automatic mode, which does not require the intervention of an expert or other external participation.

- About the child image quality in particular, the definition of quality images, significantly increases the performance of the respective software to print the Troubleshooting procedures for improving images, if not really required.

Configuring the software architecture to assess the quality of the images allows to optimize the module image enhancement by adding, removing, or arranging appropriate filter processing.

Particularly noteworthy is the fact that in the inventive solution, the metadata describing the process of correcting the image, generate and record in the file of enhanced images, which eliminates multiple quality assessment or correction of the image. This point is extremely important, because most algorithms improve digital images do not include a preliminary analysis of the image. This fact is critical because the repeated application of the correction procedure may not lead to improvement, and to the deterioration of the final image.

The invention is explained hereinafter on the basis of the following graphics:

Figure 1. A block diagram of a device that implements the inventive sorting method for automatic analysis of digital images uploaded to the printer and is stored in the memory unit, for example, the built-in hard disk of the printer.

Figure 2. Algorithm implementation the classification of digital images, correction metadata in the header Exit and sorting images before printing.

Figure 3. The classification algorithm of digital images.

Figure 4. The correction algorithm metadata in the Exif header.

Figure 5. Sorting algorithm of images.

6. The classification results.

Figure 1 shows a scheme of the interaction of the basic components of a printer that performs automatic analysis and evaluation of digital images received from the input device 101, which may be a digital camera, memory card, computer or any device that can store images containing the information, or from which digital images can be read. Digital image received from the input device 101, can be written into the memory block, for example, on the drive type hard disk, printer, or other device that can store images containing the information, or from which digital images can be read, for subsequent storage or classification. The memory block may also contain commands that are used by the software tool in the form of a module 104 for automatic analysis and correction of digital images. Assume that b is OK memory consists of two parts - program memory in which the code is running analysis, correction and printing of digital images, and data memory which records the classified image. Process analysis and image evaluation is performed by the processor 105, which executes the program stored in the block memory component 102. Information exchange is performed via the bus 103 data. After processing, the image is sent to the memory unit 102, for example, on the drive type hard disk of the printer. The metadata corresponding to the correction image, generate and write to the digital image file containing the enhanced image, to prevent multiple evaluation or correction of the image. Further data can be read from the hard disk of the printer to one or more output devices 106, which are either the print mechanism of the printer to obtain photographic prints, or any device that can store images containing the information or the device from which digital images can be read.

Figure 2 shows the algorithm of the classification process digital images for printing. In step 201 retrieves the digital image input device. In step 202 is the classification of the received images based on the analysis of brightness, analysis blur and analysis on what ice red-eye with the division into the following classes: photos to meet the high quality standards; images unsuitable for printing, but which can be improved in the future printed; images unsuitable for printing, for which improvement is possible. In step 203 is the record of the classification results in a corresponding Exif field for subsequent sorting and printing images. In step 204 high-quality image stored in the memory block, for example, drive type hard disk, printer subdirectory Normal image, which should be improved to print, recorded in the subdirectory Improvement, and images that contain fatal defects are recorded in the subdirectory Defective. In step 205 print quality images (on request).

Figure 3 shows the stages of the classification process digital images. In step 301 is classified by the brightness of the image.

To assess the quality of the exposition it is advisable to use the method described in [3], according to which for the assessment of what the image is unsuitable for printing, calculates the average brightness value of the analyzed digital image. If the average brightness is less than some threshold value I1 and is greater than some threshold value I2, the image is considered to be unsuitable for printing, the improvement is impossible. Otherwise, after the expression is classified as Normal (suitable for printing); Low contrast, Underexposed or Overexposed is unsuitable for printing, requires improvement. In case it is proposed to use the following values for 8-bit images: I1=10, I2=245.

In step 302 to produce an analysis of the degree of blur of the image, for example, by the method described in the published report, Suk Hwan Lim, Jonathan Yen and Peng Wu "Detection of Out-Of-Focus Digital Photographs", Hewlett-Packard Laboratories Technical Report HPL-2005-14 January 20, 2006 [14]. The image is assigned the rank of field from 1 to 5, where 1 is very blurred, 5 - very good. If the image has the rank of field 5, this high-quality image suitable for printing. If the evaluation of the sharpness 3-4, then this image may be improved if the sharpness is set to 1-2, then this image is defective and can not be improved.

In step 303, the program analyzes the image for the presence of red-eye. The algorithm can be roughly divided into 3 stages:

A. detection of objects that can potentially be seen as red eyes;

b. ignore false positives;

C. localize (slot) red-eye.

To determine if red-eye is recommended, in particular, to use the algorithms described in [6] and [7].

Figure 4 shows the algorithm for assigning special flag on the classification results in the Exif header of the image based on the result of the classification. At step 401, in the header of the Exif or technical image metadata boxes created and "Defect". If the image is classified as defective at step 402, the flag is "Defective" is written in the "Quality" of the Exif header in step 403. If the image is classified as "Improvement" in step 404, the flag "Improvement" is written in the "Quality" of the Exif header and the type of defect ("Defect"exposure "Defect blur" or "red-eye Defect") is written in the "Defect" in step 405. Otherwise, the flag is "Normal" is written in the "Quality" of the Exif header in step 406.

Figure 5 shows the block diagram of the sorting process is classified images in directories (folders) printer hard disk depending on the flag classification field "Quality" in the Exif header. If at step 501 is detected flag "Defective", then at step 502 for recording images in the directory "Defective", otherwise, at step 503, an analysis of the field of "Quality" in the Exif header to flag "Improvement". If the flag is set to "Improvement", in step 504 for recording images in a subdirectory of "Improvement", otherwise, at step 505, the image is saved in the subdirectory "Normal".

6 shows the classification process of the source digital images and records in the appropriate subdirectory of the memory block, for example, hard is the claim of the printer, the classification results.

As an example, at the input of the system consists of seven images, numbered from 1 to 7. The image classification results 2, 3 and 4 is defined as a quality and suitable for printing. In the "Quality" of the Exif header of the recorded flag is "Normal" and the image moved to the directory "Normal". Images 1, 5, and 7 are classified as unfit for print, because they require improvement. They in the "Quality" of the Exif header flag "Improvement", and they are recorded in the directory of "Improvement". The reason for this classification is that the image 1 is estimated as Underexposed, the image 5 is classified as Overexposed and the image 7 is classified as having red eyes. The image data must be improved before printing. Image 6 is classified as a fuzzy, without the possibility of improvement. He was given the flag is "Defective" in the "Quality" of the Exif header, and it is recorded in the directory "Defective".

The sorting method of digital images, including image quality assessment, described above, can be implemented as in printing systems, which have a hard disk and the software auto image enhancement as a plugin and as a standalone application. In most implementations of the present invention Paul the user can take advantage of automatic quality analysis and correction without the use of additional applications. It should be noted that the application of this type, classifying images on both global and local levels, makes the performance of automatic procedures higher, eliminating incorrect image enhancement, when high-quality digital images will be degraded due to incorrect classification of defects.

Thus, the present invention improves the user's ability to handle large archives of digital images by analyzing metadata associated with digital images during the execution of post-processing after shooting. Further, if there are problems with the quality of the obtained images, the user can access the metadata in the image file, in order to perform its correction. The same is true for automatic mode. In particular, the user can print only those images that were judged as "Normal", without any additional improvements. In another implementation, the user may write on a separate media only pictures from the subdirectory "Normal" to transfer them to the photo lab.

Improving the quality of images having the flag of "Improvement"is carried out depending on the type of defect ("Defect"exposure "Defect blur" or "Defect red CH is C"). If the image has a "Defect"exposure, then its correction may be performed based on the algorithm exposure correction, described in [3]. If the image has a "Defect blur", it is advisable to apply the methods of sharpening, for example, by using a number of filters, amplifying the high frequency. Different implementations of these filters are given in [13]. If the image has a red-eye defect, the correction procedure similar to that described in [7] and [8].

It is noteworthy that the inventive method can also be successfully implemented using the print device, equipped with a built-in hard drive and software for automatic evaluation of image quality that provides the user with ample opportunity to host and analyze the quality of the images, without additional software packages.

In other words, the printer is equipped with a system of assessing the quality of images will allow to organize the printing process in batch mode, even for sets is not very high quality images.

The way of sorting digital images for high-quality printing in automatic mode, comprising the following operations:

write digital images, for example, from a digital camera, a memory block that can be included in SOS the AB computer or printer,

analyze digital images contained in the memory block, on the basis of brightness of the image, the degree of blur detection red-eye, recorded in the digital image files metadata corresponding to the nature of the detected defects, classify digital images according to the results of the analysis into the following categories: images that meet the highest quality standards and is suitable for printing, i.e., without defects, images, suitable for printing, provided they are editing, i.e. with the adjustable defects, image unsuitable for printing, i.e. with fatal defects, create a block of memory sections for storing each of the three above categories digital images recorded in the memory block files are sorted digital images along with metadata about the nature of the detected defects.