JP2008148183A - Image processor, image processing method, and image processing program - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To apply metadata in accordance with background image in addition to figure image easily and correctly with respect to a digital image, to which the metadata is not given. <P>SOLUTION: Digital image is classified into a figure image and a background image (S203). Metadata in accordance with the figure image out of the metadata is applied to the digital image having the figure image, which is in agreement with the figure image, to which metadata was already given (S211). Metadata in accordance with the background out of the metadata is applied to the digital image with the background image in agreement with the background image, to which metadata was already given (S216). <P>COPYRIGHT: (C)2008,JPO&INPIT

Description

  The present invention relates to an image processing apparatus, an image processing method, and an image processing program for adding auxiliary data to a digital image having a foreground image.

  When a digital image is taken using a photographing device such as a digital camera, incidental data representing a large amount of information other than image data (eg, image resolution, exposure setting value, degree of image compression, etc.) is stored in a predetermined file. The assignment is performed according to a format (for example, “Exif (registered trademark)”). These incidental data are also called metadata. By assigning metadata to the digital image in this way, a specific digital image can be easily extracted from many digital images based on the assigned metadata.

  Further, as such metadata, metadata desired by the photographer, such as data on the location and shooting direction where the digital image was taken, and data (person name or landscape name) according to the subject can be added. It has become like this. Therefore, in the photographed digital image, for example, if the foreground image in the subject is a person, “person name” is used as metadata, and for background images other than the foreground image, “building name in the background” is used as metadata. And “photographing location” are generally given as metadata for identifying digital images.

  On the other hand, for example, there are many still pictures that have been taken and stored by a photographer with a photographing apparatus called a so-called film camera. Such still photographs are generally important for photographers who photograph people such as family members. For this reason, a photographer converts a still picture of a family member into a digital image using an input device such as a scanner, and stores the converted digital image in a predetermined recording medium such as a CD-ROM. It has become like this. The still picture is because the silver salt on the film surface deteriorates and the photographed image fades and cannot be viewed.

  Since such a still picture basically has no metadata, the digital image obtained by converting the still picture basically has no metadata. Therefore, in order to easily extract a digital image obtained by converting a still photo from many digital images stored on a recording medium such as a CD-ROM, metadata is added to the digital image converted from the still photo. It is necessary to grant.

  By the way, the work of assigning metadata to a digital image to which no metadata is attached is performed by an operator inputting appropriate data using a predetermined input means while confirming the image content of the digital image. It becomes a very troublesome work. In addition, in such an assignment operation, for example, if the name of the subject person is forgotten and the person name cannot be input as metadata, the assignment operation itself cannot be performed. Thus, Patent Document 1 proposes a method for automatically assigning metadata to a digital image to which metadata is not provided. Patent Document 1 discloses a technique for adding the name of a person in a captured image to the metadata of the image.

JP 2004-62868 A

  However, the technique disclosed in Patent Document 1 adds a person name of a known face that matches the face in the captured image as metadata, and corresponds to the background image excluding the person image that becomes the foreground image. As for metadata (for example, scenery name), no technical disclosure is made regarding the method of giving the metadata. Therefore, although it is possible to automatically assign metadata for a person image, it is necessary to perform the above-described assignment operation for a background image, and there is a problem that metadata cannot be easily assigned.

  The present invention has been made in view of such a problem, and provides an image processing apparatus, an image processing method, and an image processing program that automatically add auxiliary data to a digital image having a foreground image such as a human image, for example. For the purpose. This makes it possible to easily and correctly assign metadata corresponding to the background image in addition to the foreground image, for a digital image to which no metadata is assigned.

  An image processing apparatus according to the present invention that achieves the above object is an image processing apparatus that adds auxiliary data to a digital image having a foreground image, and stores a digital image with auxiliary data to which the auxiliary data is added. A first image that divides the digital image with ancillary data into a foreground image and a background image other than the foreground image. A classification unit, a second image segmenting unit that classifies the digital data without incidental data into the foreground image and a background image other than the foreground image, a background image segmented by the first image segmenting unit, and A coincidence determination unit that determines whether or not the background image segmented by the second image segmentation unit matches, and a digital data with ancillary data that includes the background image that is determined to be coincident by the match determination unit Issued to Le image, the supplementary data corresponding to the background image, and gist that and a supplementary data providing unit to be applied to the auxiliary data-free digital image.

  According to this configuration, the digital image is divided into a foreground image and a background image. Then, the metadata corresponding to the background among the metadata is assigned to the digital image having the background image that matches the background image to which the metadata has already been assigned.

  Therefore, since the digital image is divided into a foreground image and a background image and metadata is given, the probability that the most suitable metadata is automatically given to each image is increased. Further, since the digital image is divided into the foreground image and the background image, the influence of the foreground image on the background image can be eliminated in the background image match determination process described later, and the accuracy of the match determination can be increased. I can expect.

  Here, the foreground image may be a person image.

  In general, there are many cases where a person image has a relatively large image area shown in the entire screen area of a digital image. Therefore, it can be expected that by dividing the foreground image as a human image, the accuracy of matching determination for the divided background image is increased. Therefore, it can be expected that metadata corresponding to the background image is correctly given.

  In addition, the first image segmenting unit or the second image segmenting unit extracts a face image from the digital image having the person image, and the extracted face image region is preliminarily set according to the size of the face. The determined image area may be divided as the person image.

  In this way, since the image area determined in advance according to the size of the face is divided as a person image, the background image area and the person image area can be easily divided. Therefore, in the image matching determination process, the processing load related to the classification of person images is reduced.

  The present invention can also be understood as an image processing method. That is, an image processing method for adding auxiliary data to a digital image having a foreground image, the storing step storing the digital image with auxiliary data to which the auxiliary data is added, and the auxiliary data not being added An acquisition step of acquiring a digital image without incidental data, a first image classification step of dividing the digital image with auxiliary data into a foreground image and a background image other than the foreground image, and the digital image without incidental data A second image segmentation step of segmenting the foreground image into a background image other than the foreground image, a background image segmented by the first image segmentation step, and a background image segmented by the second image segmentation step And a match determination step for determining whether or not they match, and a digital image with accompanying data including the background image determined to match the match determination step. , The supplementary data corresponding to the background image, and summarized in that with a, and ancillary data providing step of providing the supplementary data free digital image.

  According to the image processing method of the present invention, the same operational effects as those of the above-described image processing apparatus of the present invention can be obtained. This image processing method may be executed in an image processing apparatus having the various aspects described above, or may be executed in an image processing apparatus having other aspects.

  Furthermore, the present invention can be understood as a computer program. That is, an image processing program for adding auxiliary data to a digital image having a foreground image, the storage function storing the digital image with auxiliary data to which the auxiliary data is added, and the auxiliary data not being added An acquisition function for acquiring a digital image without incidental data, a first image classification function for dividing the digital image with incidental data into a foreground image and a background image other than the foreground image, and the digital image without incidental data. A second image segmentation function for segmenting the foreground image into a background image other than the foreground image, a background image segmented by the first image segmentation function, and a background image segmented by the second image segmentation function And a coincidence determination function for determining whether or not they match, and a digital image with ancillary data including the background image determined to match the match determination function Is given, the auxiliary data corresponding to the background image, the gist incidental and data providing function, that is realized in a computer to be applied to the auxiliary data-free digital image.

  By executing this program on a predetermined operation system, the above-described image processing method is executed, and the same operational effects as those of the above-described image processing apparatus of the present invention can be obtained. This program may be recorded on a computer-readable recording medium, or may be transferred to the computer via a transmission medium such as the Internet.

  Next, embodiments of the present invention will be described based on examples. FIG. 1 is a block diagram showing an image processing apparatus 100 as an embodiment of the present invention. The image processing apparatus 100 according to the present embodiment is basically a personal computer that executes a predetermined application program under a predetermined operating system. As such a computer, not only a general-purpose personal computer as in this embodiment, but also any computer such as a server or a large computer that can run application programs can be used.

  Connected to this personal computer are a digital video camera 140, a CD-ROM (DVD-ROM) 150, a digital camera 160, and a scanner 170 as supply devices for supplying digital images such as moving images and still images. . In addition, a keyboard 131 and a mouse 132 are connected as input devices for a user operating the personal computer (hereinafter simply referred to as “user”) to perform necessary operation input. A monitor 130 is connected as a display device for displaying images and information necessary for image processing. As a digital image supply device, a USB memory or a card-type memory may be connected in addition to the one shown in FIG.

  The personal computer includes a CPU 110 that executes arithmetic processing for image processing, a ROM 111 that stores the arithmetic processing program and the like, a RAM 112 that temporarily records and stores pixel data of pixels constituting a digital image, and an image A hard disk (HD) 113 that stores application software for processing and a digital image supplied from a supply device, an I / F 120 that controls an interface with each supply device, and each input device and display device And an I / F 121 that controls the interface. They are wired so that necessary data is exchanged via the bus line 115.

  Accordingly, the CPU 110 reads out and executes the arithmetic processing program stored in the ROM 111 and the application software stored in the hard disk 113, whereby the storage unit 100a, the acquisition unit 100b, the first image classification unit 100c, and the second image. It functions as a sorting unit 100d, a first match determination unit 100e, a second match determination unit 100f, a first metadata grant unit 100g, and a second metadata grant unit 100h. Each unit mainly performs the following processing.

  The storage unit 100a stores image data of a digital image with metadata to which auxiliary data (hereinafter referred to as metadata) supplied from each supply device via the I / F 120 is added together with the assigned metadata. Store in memory. The acquisition unit 100b acquires a digital non-metadata image supplied with metadata supplied from a supply device (for example, a scanner) in a predetermined recording area of the RAM 112 and acquired. The first image classification unit 100c classifies the digital image with metadata stored in the hard disk 113 into a person image and a background image as a foreground image. The second image classification unit 100d classifies the digital image without metadata recorded in the RAM 112 into a human image and a background image as a foreground image. The first coincidence determination unit 100e compares the person image of the divided digital image with metadata and the person image of the digital image without metadata to determine whether or not the two images match. The second match determination unit 100f compares the background image of the divided digital image with metadata and the background image of the digital image without metadata to determine whether the two images match. The first metadata assigning unit 100g assigns metadata corresponding to a person among metadata attached to a digital image with metadata including a matched person image as metadata of the digital image without metadata. . The second metadata assigning unit 100h assigns metadata according to the background as metadata of the non-metadata digital image among the metadata assigned to the digital image with metadata including the matched background image. To do.

  Next, image processing performed by the image processing apparatus 100 having the above-described functional configuration unit will be described with reference to the flowchart of FIG.

  When this process is started, first, in step S201, a process of storing a plurality of digital images with metadata to which metadata has already been assigned is performed. CPU110 stores the digital image supplied from each supply apparatus via I / F120 in the hard disk 113 with metadata.

  Next, in step S202, processing for acquiring a digital image without metadata to which metadata is not assigned is performed. In this embodiment, it is assumed that the user converts a still photo into a digital image by the scanner 170 (FIG. 1). Therefore, the CPU 110 acquires the digital image obtained by converting the still picture supplied from the scanner 170 via the I / F 120 by recording the image data in the RAM 112.

  FIG. 3A shows an example of a digital image obtained by converting a still photograph supplied from the scanner 170. As described above, this digital image is a digital image without metadata, and is a person image taken with a mountain as a background. In this embodiment, as metadata to be added to the digital image, metadata corresponding to a person (for example, a person name) and metadata corresponding to a background (for example, a background name) are already added. It is intended to obtain and add from a digital image with metadata.

  Next, in step S203, processing for dividing the digital image without metadata into a human image and a background image is performed. By this processing, as shown in FIG. 3B, a person image without a background image and a background image from which a person image (broken line in the figure) is cut out as shown in FIG. It is divided into.

  By the way, in this embodiment, the person image is extracted from the digital image, and the person image and the background image are classified. The CPU 110 reads application software for the person image extraction process stored in the hard disk 113 and performs a predetermined process to extract the person image. For example, a face image is detected by skin color detection processing, an image region corresponding to a person's body is searched by region division processing, and a boundary between the body and the background image is searched by edge extraction processing. In addition to these processing techniques, many processing techniques such as a technique for detecting a nostril and detecting a face image have already been disclosed as person image extraction processing. In this embodiment, these processing techniques are disclosed. It is assumed that a person image is extracted using a technique.

  Next, in step S205, a process of reading one of the digital images with metadata is performed. The CPU 110 reads one digital image with metadata stored in the hard disk 113 in a predetermined order (for example, in chronological order from the oldest shooting date / time). In step S206, the digital image with metadata is divided into a person image and a background image.

  An example of the digital image with metadata read out in step S205 is shown in FIG. This digital image is a human image taken with “Δ lodge” in the background. Then, it is assumed that the person name “Taro” and the background name “Δ Lodge” are given to the digital image as metadata.

  In step S206, the digital image with metadata is divided into a person image and a background image, whereby a person image having no background image as shown in FIG. ), The person image (broken line in the figure) is divided into the cut-out background image. The method for classifying the person image and the background image is performed in the same manner as the method described in step S203 described above. Therefore, the description is omitted here.

  Next, in step S207, a process for comparing person images and background images is performed. In this embodiment, image patterns and image data histograms are used, human images are compared by image data pattern matching, and background images are compared by image data histogram comparison. Note that pattern matching and histograms are well-known processing techniques in the image processing technique, and thus descriptions of these processes are omitted.

  Here, in the comparison process between human images, in this embodiment, pattern matching is performed on the detected face image. The CPU 110 extracts a skin color image area from the person image, and performs pattern matching processing using the extracted image area as a face image. In the pattern matching process, the size of one of the face images is enlarged or reduced to a size that approximately matches the size of the other face image so that the two face images to be compared have the same size. To do. The image enlargement or reduction processing can be performed by an interpolation processing method (for example, a bi-linear method) of pixels constituting the image.

  Next, in step S210, it is determined whether or not the person images match. In this embodiment, the CPU 110 compares the pixel values of the pixels constituting the two face images, and determines whether the total sum of the pixel value differences for the corresponding pixels is within a predetermined threshold value. It is determined whether the face images match. Of course, any other method may be used as long as it is a method capable of determining the coincidence of face images (for example, the method disclosed in Patent Document 1).

  As a result of the determination, if the face images match, that is, the person images match (step S210: YES), in step S211, processing is performed to add metadata according to the person of the read digital image with metadata. The process proceeds to S215. In the present embodiment, since the person image of the digital image shown in FIG. 3B matches the person image of the digital image shown in FIG. 4B, the CPU 110 shows that shown in FIG. Among the metadata of the digital image with metadata, “Taro”, which is metadata corresponding to the person, is given as metadata of the digital image without metadata shown in FIG.

  On the other hand, if the person images do not match as a result of the determination (step S210: NO), the process proceeds to the next step S215 without doing anything.

  Next, in step S215, it is determined whether or not the background images match. In the present embodiment, the CPU 110 represents the two background images as histograms indicating the relationship between the gradation value of the pixel and the appearance frequency as described above, compares the two represented histograms, and determines the respective gradation values. It is determined whether or not the two images match depending on whether or not the appearance frequency is within a predetermined frequency difference. Of course, any other method may be used as long as it can determine whether the background image matches.

  As described above, in this embodiment, the person image and the background image are classified, and an image obtained by removing the person image from the digital image is used as the background image. Accordingly, since the influence of the human image is almost eliminated from the represented histogram, only the background image can be correctly compared.

  If the background images match as a result of the determination (step S215: YES), in step S216, processing is performed to add metadata according to the background of the read digital image with metadata, and the process proceeds to step S220.

  On the other hand, if the background images do not match as a result of the determination (step S215: NO), the process proceeds to the next step S220 without doing anything. In the present embodiment, the background image of the digital image without metadata shown in FIG. 3C and the background image of the digital image with metadata shown in FIG. No metadata is given.

  In step S220, it is determined whether or not there is a digital image with metadata that has not been read out. If it exists as a result of the determination (YES), the process returns to step S205 to repeat the processing up to step S220 described above. The CPU 110 reads out the next digital image in accordance with a predetermined order from the digital images that have not yet been read out with respect to the digital images with metadata stored in the hard disk 113.

  An example of the read digital image with metadata is shown in FIG. This digital image is a human image taken against the background of “mountain and mountain”. It is assumed that the name of the person “Jiro” and the background name “Mt.

  In step S206, a process for dividing the image into a person image and a background image is performed. The digital image with metadata includes a person image shown in FIG. 5B and a background image shown in FIG. It is divided into. The method for dividing the person image and the background image is the same as the method described in step S203.

  Next, as described above, in step S207, person images and background images are compared, and in step S210, a determination process is performed to determine whether the person images match.

  Here, since the person image shown in FIG. 3B and the person image shown in FIG. 5B do not match (step S210: NO), the process proceeds to the next step S215 without doing anything.

  Next, in step S215, a determination process is performed to determine whether the background images match. If the background images match (step S215: YES), in step S216, processing is performed to add metadata according to the background of the read digital image with metadata, and the process proceeds to step S220. This time, the background image of the digital image without metadata shown in FIG. 3C matches the background image of the digital image with metadata shown in FIG. Among the metadata of the digital image with metadata shown in FIG. 3), “uneven mountain” that is metadata according to the background is given as metadata of the digital image without metadata shown in FIG.

  Such a process is performed for all digital images with metadata stored in the hard disk 113, and when there are no metadata-ready digital images that have not been read (step S220: NO), the image in this embodiment is displayed. The image processing in the processing apparatus 100 ends.

  Incidentally, in this embodiment, the metadata-less digital image shown in FIG. 3A is given “Taro” and “Roughness mountain” as metadata as a result of the above-described metadata assignment processing. Therefore, if the digital image that the user wants to extract is an image of “Taro” against “Mt. Ridge”, the user can easily obtain the desired digital image based on the metadata “Taro” and “Mt. Can be extracted.

  As described above, according to this embodiment, a digital image to which no metadata is added is divided into a person image and a background image. Then, among the metadata already assigned to the digital image in which the person image that matches the divided person image is included, metadata corresponding to the person is assigned to the digital image to which no metadata is assigned. Further, among the metadata already added to the digital image in which the background image matching the segmented background image is reflected, the metadata corresponding to the background is assigned to the digital image to which no metadata is assigned. As a result, a digital image that the user desires to extract can be easily extracted from a large amount of digital images using metadata.

  Further, since the metadata is given by dividing the person image and the background image, the probability that the most suitable metadata is given to each image is increased. Further, since the person image and the background image are separated, it is possible to eliminate the influence of the person image on the background image in the background image matching determination process, and it is expected that the accuracy of the matching determination for the background image will be improved. it can.

  The embodiments of the present invention have been described with reference to the examples. However, the present invention is not limited to these embodiments, and can be implemented in various forms without departing from the spirit of the present invention. Of course. Hereinafter, a modification will be described.

(First modification)
In the above embodiment, in step S203 (and step S206) in FIG. 2, in addition to the face image detection process, a person image is extracted by performing image processing such as region division processing and edge extraction processing. However, such a process requires a large number of calculations and requires a large memory area to be used, so that the load according to the image processing becomes heavy. Therefore, as a modified example, the person image may be an image region determined in advance based on the extracted face image. By doing so, it is possible to reduce the load according to the image processing, and it can be expected that the processing speed is increased.

  This modification will be described with reference to FIGS. 6 and 7. FIG. 6 is a schematic diagram illustrating an example of setting the image area of the person image with respect to the extracted face image. As shown in the figure, the person image area has a length KW1 on the left side in the drawing from the line KCX which is the center position of the width KW of the face image in the horizontal direction (left and right direction in the drawing) as shown by the shaded portion in the drawing. The right side of the drawing is the length KW2, and the vertical direction (the vertical direction of the drawing) is the length KH1 on the upper side of the drawing and the length KH2 on the lower side of the drawing from the line KCY that is the center position of the vertical width KH of the face image.

  The length KW1 and the length KW2 are obtained by setting the length assuming the shoulder width with respect to the lateral width KW of the face in the lateral direction with the face as the center. The length KH1 and the length KH2 are set with respect to the length KH of the face, assuming the length to the hair portion and the length to the feet. Thus, if the image area of the person image is determined in advance, the image area of the person image can be roughly extracted by detecting the face image.

  An example of the case where the image area of the person image is extracted in this way is shown in FIG. FIG. 7A shows a case where the person image is divided by the dividing method according to this modification example with respect to the digital image without metadata shown in FIG. 3A, and FIG. The figure shows a case where a person image is segmented by the segmentation method according to this modification to the digital image with metadata shown in (a).

  In FIG. 7, a rectangular portion indicated by a shaded area in the drawing is a portion of a divided person image. As is clear from the figure, in the metadata-free digital image shown in FIG. 7A, a part of the background image is classified as a person image, but a part of the person image is hardly classified as a background image. The probability that a person image will have an effect on the background image matching process is small. In addition, in the digital image with metadata shown in FIG. 7B, a part of the human image (left and right hands) is classified as a background image, but the image area is smaller than the size of the background image. Since the histogram does not change greatly, it can be considered that the probability that the human image has an influence in the matching determination processing is small.

(Second modification)
In the above-described embodiment, pattern matching is performed on the detected face image in the comparison process between human images in step S207 of FIG. In the present modification, the entire human image may be compared using a histogram.

  Thus, by comparing the whole person images, it is effective when it is difficult to perform pattern matching of face images because, for example, a person appears sideways or a face appears small. That is, since the entire person images are compared, for example, a digital image with matching clothes can be extracted. As a result, it is possible to extract a digital image with matching clothes, that is, with a matching person, and to correctly assign metadata corresponding to the person. Further, since human images can be compared without performing facial image enlargement / reduction processing, the processing load may be reduced.

  As a specific comparison processing method, similar to the background image comparison processing described above, the two person images are represented as histograms showing the relationship between the gradation value of the pixel and the appearance frequency, and the two represented histograms are expressed as follows. It is only necessary to compare and determine whether the two images match depending on whether the appearance frequency of each gradation value is within a predetermined frequency difference. Needless to say, any other method may be used as long as it can determine the matching of person images.

(Third Modification)
In the above embodiment, all the digital images with metadata stored in the hard disk 113 are read one by one in a predetermined order and compared with the digital images without metadata. This is because, although not described above, it is possible to assign all metadata according to the person and all metadata according to the background for the metadata given in advance. .

  However, when the pre-assigned metadata is the same or approximately the same, it is not always necessary to compare the digital image with metadata with all the digital images. Therefore, as a modified example, there is a digital image that matches a person image, and when metadata corresponding to a person previously assigned to the digital image is added to the digital image without metadata, The comparison process may be terminated. Similarly, there is a matching digital image for the background image, and when the metadata corresponding to the background previously given to the digital image is given to the digital image without metadata, the comparison process for the background image is performed. It is good also as ending. In this way, the load corresponding to the image processing is reduced.

  Although the specific processing in this modification is not shown, when the processing in step S211 is performed once in the processing flowchart shown in FIG. 2, the processing in step S210 and step S211 is not performed thereafter. . Further, when the process of step S216 is performed once, the processes of step S215 and step S216 are not performed thereafter. Then, all the processes may be terminated when both the processes of step S211 and step S216 are performed once.

(Other variations)
In the above embodiment, the digital image without metadata is acquired from the scanner 170, but the present invention is not limited to this. For example, a still picture may be taken with the digital video camera 140 or the digital camera 160, and the taken digital image may be acquired. Or it is good also as acquiring from CD-ROM150 in which the image which metadata, such as a still picture, is not provided is recorded as a digital image.

  When there are many digital images to be acquired at the time of acquisition, a list of the digital images to be acquired may be displayed on the monitor 130 shown in FIG. 1, and the user may select the digital images by operating the keyboard 131 or the mouse 132. .

  Further, the digital image without metadata may be a digital image converted from a moving image in addition to a digital image converted from a still image such as a still photo. In this case, a frame image constituting a moving image is supplied from the digital video camera 140 to the image processing apparatus 100. When the digital image without metadata is a frame image of a moving image, if a person who is a subject is moving, as is well known, the person image is analyzed by motion vector analysis between consecutive frame images in time series. Can be extracted.

  In the above embodiment, all the digital images with metadata that match the person image or the background image are displayed in a list on the monitor 130, and the user can select the digital images with metadata to which the metadata should be added. The selection may be made by operating the keyboard 131 or the mouse 132. In this way, metadata desired by the user can be appropriately given.

  In the above embodiment, the metadata is assigned to both the person image and the background image. However, the metadata may be assigned to only the person image or only the background image. In the above embodiment, a human image is targeted as a foreground image. However, the present invention is not limited to a human image and can be applied to a specific image as a foreground image.

1 is a configuration diagram of an image processing apparatus as an embodiment of the present invention. 6 is a flowchart for explaining processing performed by the image processing apparatus according to the embodiment. (A) is a schematic diagram which shows an example of the digital image which converted the still photograph. (B) is a schematic diagram showing a person image segmented from the digital image. (C) is a schematic diagram showing a background image segmented from the digital image. (A) is a schematic diagram which shows an example of the read digital image with metadata. (B) is a schematic diagram showing a person image segmented from the digital image. (C) is a schematic diagram showing a background image segmented from the digital image. (A) is a schematic diagram which shows an example of another read digital image with metadata. (B) is a schematic diagram showing a person image segmented from the digital image. (C) is a schematic diagram showing a background image segmented from the digital image. The schematic diagram which showed a mode that the image area of the person image was defined with respect to the face image. (A) is explanatory drawing which shows the case where a person image is divided by the classification method by the 1st modification. (B) is explanatory drawing which shows the case where a person image is divided by the classification method by the 1st modification.

Explanation of symbols

  DESCRIPTION OF SYMBOLS 100 ... Image processing apparatus, 100a ... Memory | storage part, 100b ... Acquisition part, 100c ... 1st image division part, 100d ... 2nd image division part, 100e ... 1st coincidence determination part, 100f ... 2nd coincidence determination , 100 g... First metadata assigning unit, 100 h... Second metadata assigning unit, 110... CPU, 111... ROM, 112... RAM, 113. ... I / F, 130 ... monitor, 131 ... keyboard, 132 ... mouse, 140 ... digital video camera, 160 ... digital camera, 170 ... scanner.

Claims (5)

An image processing apparatus for attaching auxiliary data to a digital image having a foreground image,
A storage unit for storing a digital image with accompanying data to which the accompanying data is attached;
An acquisition unit for acquiring a digital image without incidental data to which the incidental data is not attached;
A first image segmentation unit for segmenting the digital image with accompanying data into the foreground image and a background image other than the foreground image;
A second image segmentation unit that segments the digital data without incidental data into the foreground image and a background image other than the foreground image;
A coincidence determination unit that determines whether or not the background image segmented by the first image segmentation unit and the background image segmented by the second image segmentation unit match;
An auxiliary data giving unit that is attached to the digital image with accompanying data that includes the background image determined to be matched by the matching determining unit, and that attaches the auxiliary data according to the background image to the digital image without accompanying data;
An image processing apparatus. The image processing apparatus according to claim 1,
The image processing apparatus according to claim 1, wherein the foreground image is a person image. The image processing apparatus according to claim 2,
The first image segmenting unit or the second image segmenting unit extracts a face image from the digital image having the person image, and the extracted face image region is predetermined according to the size of the face. An image processing apparatus, wherein the image area is divided as the person image. An image processing method for attaching auxiliary data to a digital image having a foreground image,
A storage step of storing a digital image with accompanying data to which the accompanying data is attached;
An acquisition step of acquiring a digital image without incidental data to which the incidental data is not attached,
A first image segmentation step of segmenting the digital data with accompanying data into the foreground image and a background image other than the foreground image;
A second image segmentation step of segmenting the auxiliary data-free digital image into the foreground image and a background image other than the foreground image;
A matching determination step for determining whether the background image divided by the first image dividing step and the background image divided by the second image dividing step match,
Ancillary data providing step of attaching the incidental data according to the background image to the digital image with the incidental data, which is given to the digital image with the incidental data including the background image determined to match the coincidence determining step;
An image processing method comprising: An image processing program for adding auxiliary data to a digital image having a foreground image,
A storage function for storing a digital image with accompanying data to which the accompanying data is attached;
An acquisition function for acquiring a digital image without incidental data to which the incidental data is not attached;
A first image segmentation function for segmenting the digital image with accompanying data into the foreground image and a background image other than the foreground image;
A second image segmentation function for segmenting the auxiliary data-free digital image into the foreground image and a background image other than the foreground image;
A matching determination function for determining whether or not the background image segmented by the first image segmentation function matches the background image segmented by the second image segmentation function;
Attached data giving function that is given to the digital image with accompanying data that includes the background image determined to match the matching judgment function, and that gives the accompanying data according to the background image to the digital image without accompanying data;
An image processing program for realizing a computer.

Images