JP2006018523A - Image processing device, image processing method, and computer program - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To sequentially designate a detection parameter for setting the size or application range of a template used when a specific object is detected in a plurality of images without a troublesome operation and spending much time in detection processing. <P>SOLUTION: When a detection parameter is determined to change based on a correlation between a previous image and an image to be processed (S106), a face detection parameter is changed (S107). In this parameter changing processing, a face is adequately detected with a detection parameter even if the range of the detection parameter is limited to a smaller range in case that the previously processed image has a high correlation with the image to be processed. By this, when a detection object such as a face is detected by moving or changing a template according to the parameter, the range for the movement and change can be reduced, and time required for detection processing can also be decreased. <P>COPYRIGHT: (C)2006,JPO&NCIPI

Description

  The present invention relates to an image processing apparatus, an image processing method, and a computer program, and relates to processing for detecting a specific target such as a human face in image data obtained by photographing with a digital camera or the like.

  In recent years, digitization of photographic images has become easier due to the spread of digital cameras. In addition, the quality of printing apparatuses, particularly inkjet printing apparatuses, has increased, and it has become possible to obtain photographic print results from digital images obtained by users. Furthermore, it is also easy to edit and process these captured image data using various application software on a personal computer.

  In view of the above background, development of an environment in which a photographed image can be printed easily and with high quality is being promoted. As one of such applications, an application for detecting a human face from image data obtained by photographing and performing appropriate image processing on the human face has been developed. This is because a person (face) is often the center of a snapshot taken by a digital camera, and image data is corrected for the detected face, for example, a more natural facial impression. This is because printing is performed. For this reason, many methods for detecting a face from an image have been proposed.

  As one of methods for detecting such a face, a method using a template is known. In Patent Document 1, if a matching degree between a template having a plurality of face shapes and an image is calculated and the matching degree is the highest, if the matching degree is equal to or higher than a predetermined threshold, an area in the template is disclosed. Is a face candidate region. It is also described that the position of the eye can be detected here and here using this template.

  Similarly, the one described in Patent Document 2 is known as one using a template. Here, first, the nose image pattern is used as a template, the entire image or a designated area in the image is scanned, and the best matching position is output as the nose position. Next, the region above the nose position of the image is considered as the region where the eyes exist, and the eye presence region is scanned and matched using the eye image pattern as a template, and a set of pixels having a degree of matching greater than a certain threshold An eye presence candidate position set is obtained. Further, continuous regions included in the eye presence candidate position set are divided as clusters, and the distance between each cluster and the nose position is calculated. It is described that each organ of the face is detected by determining the cluster having the shortest distance as the cluster in which the eye exists.

  In the technology for detecting a face using the template as described above, a plurality of templates having different sizes, orientations such as upside down or sideways of the image of the face or the like in the photographed image or the nose and eyes are used. A matching process is performed by using them and selectively using them. In this way, in a technique for detecting a specific target by setting a detection area having a certain size, not limited to a template, and applying it to the image, the size and direction of the area or the movement range thereof is used in the image. The range of detection processing when detecting a detection target is set as a detection parameter, and the above-described region is set by designating those parameters.

  Regarding the detection of the face and organ positions, many other methods such as Patent Documents 3 to 12 have been proposed.

JP-A-8-63597 JP 2000-105829 A JP 2002-183731 A JP 2003-30667 A JP-A-8-77334 JP 2001-216515 A JP-A-5-197793 JP-A-11-53525 JP 2000-132688 A JP 2000-235648 A JP-A-11-250267 Japanese Patent No. 2541688

  However, when a specific target is detected by sequentially applying a detection area such as a template for a plurality of images obtained by shooting or the like, setting of the detection area for each image is troublesome or the detection area is used. There is a problem that a relatively long time is required for the detection process.

  That is, when the user designates a detection parameter according to the content of each image, the operation for that is complicated. For example, when detecting a face from an image, the maximum and minimum values of the size of the detected face, the number of detected faces, the direction of the top and bottom of the face, etc. are used as detection parameters. For this reason, each of these parameters must be specified in comparison with the contents of the target image, which is a complicated operation.

  On the other hand, a configuration in which detection parameters are specified without user operation is conceivable. In this case, generally, a general-purpose detection parameter is determined in advance so as to cope with various face sizes and directions in the captured image, and detection is performed using this as a default detection parameter. I am doing so. However, in this case, since the default detection parameters are those for which versatility is obtained, it is required to perform strict detection in the detection using these parameters. That is, in order to avoid limiting the size and movement range of the detection target in such detection, all or a relatively large number of detection parameters are specified, and detections set by the specified detection parameters, respectively. The detection is performed using the work area. This makes it possible to accurately detect a detection target such as a face. However, when performing such exact detection, a lot of time is required, and further, since such exact detection is performed for a plurality of images, a much longer time is required.

  The present invention has been made to solve such a problem, and the object of the present invention is to detect the size of a detection region such as a template used when detecting a specific target in a plurality of images and the size thereof. To provide an image processing apparatus and an image processing method that do not require a complicated operation and that do not require much time for detection processing when specifying detection parameters for setting an application range or the like sequentially. It is in.

  Therefore, the present invention provides an image processing apparatus that detects a predetermined detection target in an image using a detection parameter that defines a detection processing range when the detection target is detected in the image, and detects the predetermined detection target. Correlation judging means for judging the degree of correlation between the image subjected to the detection and the image for detecting the predetermined detection object next to the image, and the next judgment according to the degree of correlation judged by the correlation judging means And a correction means for correcting a detection parameter used for the detection in an image for detecting a predetermined detection target.

  An image processing method for detecting a predetermined detection target in an image using a detection parameter that defines a detection process range when detecting the detection target in the image, wherein the predetermined detection target is detected. A correlation determination step for determining the degree of correlation between the broken image and the image for which the predetermined detection target is detected next to the image, and the predetermined predetermined number according to the degree of correlation determined by the correlation determination unit. And a correction step of correcting a detection parameter used for the detection in an image for detecting a detection target.

  According to the above configuration, the degree of correlation between the image where the predetermined detection target is detected and the image where the predetermined detection target is detected next to the image is determined, and the degree of correlation determined Then, the detection parameter used for detection of the image for performing the detection of the predetermined detection target next, which corrects the detection parameter that defines the detection processing range when the detection target is detected in the image, is corrected. In the case of an image having a high correlation with the image related to the detection of the detection target, the detection target can be appropriately detected even if the detection processing range is narrowed. And since the range of detection processing can be narrowed, the time required for detection can be reduced.

  As a result, in a plurality of images, it is necessary to perform complicated operations when sequentially specifying detection parameters for setting the size of a detection region such as a template used when detecting a specific target and its application range. In addition, it is possible to shorten the time for the detection process using the designated detection parameter.

  Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings.

The image processing apparatus Figure 1 is a block diagram particularly showing a hardware configuration of a printing system according to an embodiment of the present invention. In the present embodiment, as described below, the present invention is applied to a printing system that prints a photographed image with an inkjet printer based on image data obtained by photographing with a digital camera. However, the application of the present invention is not limited to such a printing system, and any system can be used as long as it detects a specific target such as a face from an image captured using a detection area specified by a detection parameter. The present invention can be applied to the following description.

  In FIG. 1, an image processing apparatus 101 is implemented in the form of a personal computer. The image processing apparatus 101 includes a CPU 102, a ROM (Read Only Memory) 103, a RAM (Random Access Memory) 104, a secondary storage device 105, a display device 106, an input device 107, an I / F control unit 108, and the like.

  The CPU 102 is a central processing unit, and in accordance with the programs stored in the ROM 103, the RAM 104, and the secondary storage device 105, the data processing in this apparatus including the image processing described in FIG. Execute control. The ROM 103 stores a program called BIOS (BasicInput / OutputSystem) for controlling input / output of a peripheral device, and is executed when the power is turned on to initialize the RAM 104 and other input / output devices. The RAM 104 provides a working memory space. The secondary storage device 105 includes a large-capacity magnetic storage device represented by an HDD (Hard Disk Drive) and a CD-ROM (Compact Disk-ROM), and stores an OS (Operating System), applications, and various data.

  The display device 106 is a display device such as a CRT (Cathode Ray Tube) display or a liquid crystal display. The display device 106 displays application processing results and messages. The input device 107 is a device for inputting user instructions such as a keyboard and a mouse, and provides a function of inputting user instructions to the image processing apparatus 101.

  The I / F control unit 108 provides an interface with an external peripheral device. As an interface, a serial bus such as IEEE1284, which is also called a parallel port, SCSI (Small Computer System Interface), USB (Universal Serial Bus), IEEE 1394, or the like can be used. Which interface is used is determined by the configuration of the image processing apparatus 101 and the correspondence of external peripheral devices.

  A digital still camera 109, a memory card reader / writer 110, and a printer 111 are connected as external peripheral devices. The digital still camera 109 is a photographing device and stores photographed image data photographed by a user. Then, the captured image data is attached to the memory card reader / writer 110 by connecting the digital still camera 109 to the image processing apparatus 101 via the I / F control unit 108 or by removing the recording medium built in the digital still camera. By doing so, it is transferred to the secondary storage device 105 and stored. The stored image data is subjected to image processing including face detection described in FIG. Furthermore, printing can be performed by the printer 111 based on the image data subjected to this image processing.

Image Format FIG. 2 is a diagram showing an example of a format 201 of image data obtained by photographing with the digital still camera 109. This format is a format when stored in the recording medium of the digital still camera 109 or a format when transferred to the image processing apparatus 101.

  The format 201 is roughly composed of a shooting information area 202 and a shot image data area 203. Generally, since the data amount of the photographing information area 202 is not so large, it is in the form of uncompressed data, and the photographed image data area 203 is in a compressed form because of the large amount of data, and has a hybrid configuration. However, the present invention is not limited to the data compression / non-compression configuration, and the photographed image data may be compressed even if the photographed image data is uncompressed.

  The shooting information area 202 includes various shooting conditions when shooting with the digital still camera 109, such as the number of vertical / horizontal pixels, exposure conditions, presence / absence of strobe issuance, white balance conditions, shooting mode, and shooting time. Stores shooting information. The shooting information data includes an ID number, a data format, a data length, an offset value, and shooting information specific data corresponding to the shooting information. As the format, for example, Exif (Exchangeable Image Format) defined by JEIDA (Japan Electronic Development Association) can be used.

  The captured image data area 203 can be captured image data of RGB data or YCC (luminance color difference) data. For example, if the image data in the captured image data area 203 is YCC data, the image processing apparatus 101 converts the YCC data to RGB data and performs image processing.

(First embodiment)
Image processing including detection of a human face in captured image data in the image processing apparatus according to an embodiment of the present invention described above will be described below.

  FIG. 3 is a flowchart showing image processing related to face detection accompanied by detection parameter change processing for designating a detection region according to the first embodiment of the present invention. When this process is executed, a plurality of photographed image data photographed by the digital still camera 109 is already stored in the secondary storage device 105 of the image processing apparatus 101.

  In FIG. 3, first, in step S101, a list of photographed image data is displayed on the display device 106, and target images to be subjected to image processing are selected according to a selection input by the user via the input device 107, and these are selected. Create a list of image data.

  In step S102, the image data at the top of the image data list created in step S101 is set as a detection target image, and a default detection parameter and a detection parameter change flag “0” (not changed) are set. initialize.

  In the present embodiment, as will be described later with respect to the processing after step S106, it is determined that the detection parameter is changed or not changed according to the degree of correlation between a plurality of image contents to be sequentially processed. As the detection parameter, a default detection parameter is set as an initial value. The detection parameter defines the range of the detection process including the size, position or direction, number, etc., of detecting a detection target such as a face in the image. In the present embodiment, parameters such as face detection size, detection area, number of faces detected in one image, and face direction are used. As for the size of face detection, the vertical and horizontal sizes are set to the maximum number of pixels on the short side of the image, and a predetermined ratio (for example, 5%) of the number of short side pixels is set to the minimum value. It takes a value. The maximum detection area is a square area with the short side of the image as one side, and detection is performed while moving the square area in the long side direction in a predetermined step. In addition, the maximum number of face detections is 256, and the direction of the face is maximum in all directions of the top and bottom directions and the left and right directions orthogonal to the top and bottom directions. The default detection parameter is set with the above-mentioned maximum parameter.

  After initialization, the detection target image is read from the secondary storage device 105 in step S103. The read captured image data has the format 201 shown in FIG. 2, the captured information area 202 is uncompressed, and the captured image data area 203 is JPEG (Joint Photographic Expert Group) compressed.

  The photographing information area 202 is analyzed in step S104, while the photographed image data area 203 is subjected to JPEG decoding processing in step S105, returned to uncompressed image data, and further changed from the YCC format to the RGB format. The color conversion process is performed (step S105).

Capturing information analysis Figure 4 is a flowchart showing the photographing information analysis processing in step S104. First, in step S201, it is determined whether shooting information is added to the shot image data being processed at that time. In general, shooting information is added to shot image data shot by a digital still camera, but this determination is performed because the shooting information may be deleted when editing or processing is performed by an application. If there is shooting information, the process proceeds to step S203. If there is no shooting information, the process proceeds to an end process in step S207, and the shooting information analysis process is ended as a shooting information error.

  In step S202, a memory area having a capacity for storing necessary photographing information is secured. If it can be secured, the process proceeds to step S203. If it cannot be secured, the process proceeds to step S207, and similarly, the photographing information analysis process is terminated as a memory securing error.

  In step S203, the shooting information is scanned item by item, and the ID value of the shooting information is searched. If the ID is detected, the process proceeds to step S204. In step S204, the detected ID is compared with the ID in the photographing information management table. Only the necessary shooting information ID is stored in the shooting information management table. Here, the imaging information # 0 and # 1 will be described as information necessary for future processing without particularly limiting the imaging information.

  Specifically, when one ID is scanned in step S203, it is confirmed in step S204 whether the ID is obtained by comparison with the list of IDs in the photographing information management table. If the ID is an ID indicating shooting information # 0, the process proceeds to step S205, where the contents of shooting information # 0 are read and the contents are copied to the memory area stored in step S202. When the copying is completed, the process proceeds to S203 for scanning the next ID. Similarly, when the ID is scanned in step S203 and the IDs are compared in step S204, and the ID is ID indicating shooting information # 1, the process proceeds to step S206, and the content of shooting information # 1 is read in step S202. Copy the contents to the stored memory area. Further, the IDs are compared in step S204, and if it is determined that the shooting information indicated by the ID is not necessary, the process proceeds to step S203 in order to execute the next ID scan without storing the contents in the memory. In this way, the retrieval and storage of the photographing information is repeated, and when the retrieval of all the IDs is completed, the process proceeds to step S207, and the photographing information analysis process is terminated as normal completion.

  Note that not all of the expected shooting information can be obtained in this shooting information analysis process. This is because the types of shooting information that are supported vary depending on the model of the digital still camera. For this reason, the subsequent processing is executed using the collected imaging information.

Detection Parameter Change Determination FIG. 5 is a flowchart showing a process for determining whether or not the face detection parameter needs to be changed in step S106 shown in FIG.

  If the target image is the first image in the processed image list created in step S101 shown in FIG. 3, the detection parameter change flag is set to “0” in step S102 in FIG. In this case, in step S301, since it is the first image, it is determined that there is no result of the previous image, the process proceeds to step S305, the detection parameter change flag remains “0”, and this determination process ends.

  If the target image is the second or subsequent image in the processed image list, it is determined in step S301 that there is a previous result, and the process proceeds to step S302. In step S302, the shooting information of the previous image and the shooting information of the image to be processed are compared, and the degree of correlation between them is examined.

  FIG. 6 illustrates various combinations of evaluation amounts such as various shooting information such as shooting modes and shooting times, histograms of pixel values of image data, and combinations of the respective evaluation amounts for determining the correlation between two adjacent images. It is a figure explaining how to judge.

  As shown in FIG. 6, in the present embodiment, three types of determinations 1 to 3 are sequentially performed according to the combination of evaluation amounts, and the sum of evaluation amounts is higher than a threshold value determined for each determination in any determination. When it is judged that the correlation is high. The determinations 4 to 6 can be used in place of other examples of the determination 3 as will be described later.

  In the above, when the evaluation amount is added, each evaluation amount is added with the weight described in parentheses in FIG. Each evaluation amount is normalized to 1, added to the evaluation amount with a weight, and if the result is equal to or greater than a predetermined threshold, it is determined that the correlation is high.

  In judgment 1 shown in FIG. 6, it is checked whether or not the shooting mode is the continuous shooting mode. In the continuous shooting mode, the images before and after are likely to be highly correlated images. In addition, the possibility of the continuous shooting mode also increases when the focal lengths are the same in the mutual images. In this determination 1, in the continuous shooting mode, the shooting time and other evaluation quantities cannot be examined. In this determination, the evaluation amount of the shooting mode has a weight of 0.7, and the evaluation amount of the focal length has a weight of 0.3. This is because the shooting mode indicates whether or not direct continuous shooting is used, and thus the weight is increased.

  In decision 2, the shooting time is examined. When the shooting time, which is the time from the previous image to the target image, is within a predetermined time (for example, within 1 minute), the same image, that is, a highly correlated image was taken. Judge that the possibility is high. For example, the same shooting may be performed preliminarily with a group photo or the like, and a plurality of images may be shot within a certain time. In addition, if the focal length (shown in FIG. 2) of the shooting information is the same as the previous image, the possibility that the image is similar is further increased. Further, in the case of continuous shooting, the shooting time is within a predetermined time, and therefore, it is adopted as an evaluation amount. However, in this determination, the weight is reduced.

  In decision 3, the histogram of the photographic image data itself, not the photographic information, is analyzed and used as the evaluation amount. For the two images to be compared, a histogram of density values for each color of RGB of the image data is created, and the absolute values of the differences between the density values are added together, and the sum (error) of the differences is below a predetermined threshold value. In some cases, it is determined that the two images are the same.

  Note that in the determination 4, the distribution shape of the histogram is adopted as the evaluation amount, and when these distributions are compared, two images are determined to be the same if they have similar distributions. This determination 4 may be performed instead of the above determination 3. Further, as in determination 5, the two comparison results of determinations 3 and 4 may be weighted and averaged with a predetermined value, and the degree of correlation between the two images may be determined from the value. Further, as in decision 6, some combinations of evaluation amounts may be used. Needless to say, an amount that can be calculated from shooting information and image data other than the evaluation amount shown in FIG. 6 may be used as the evaluation amount.

  Referring to FIG. 5 again, based on the comparison result between the evaluation amount and the threshold value in step S302 described above, in the next step S303, it is determined whether the correlation between successive images is high. If it is determined that the correlation is high, the detection parameter change flag is set to “1” (change) in step S304. On the other hand, if it is determined that the correlation is not high, the detection parameter change flag is set to “0” (not changed) in step S305, and the detection parameter change determination process in step S106 is terminated.

  Referring to FIG. 3 again, when the detection parameter is not changed in step S106, that is, when the detection parameter change flag is set to “0”, the process proceeds to step S108. On the other hand, when the detection parameter is changed in step S106, that is, when the detection parameter change flag is set to “1”, the process proceeds to step S107, and the face detection parameter is changed.

  In this change, assuming that the correlation with the previously processed image is high, the face detection parameter used in the image to be processed is determined based on the face detection result of the previously processed image. That is, in this parameter changing process, when the correlation with the previously processed image is high, even if the range of the detection parameter to be used is limited to a smaller range, the face can be detected appropriately with that parameter. A process for narrowing the detection parameter range is performed.

  For example, when the face is photographed large like a portrait photograph in the previous image and a large area is detected with respect to the image size as a face detection result, the size of the face to be detected is defined. In order to narrow the detection parameter range, for example, the minimum value of the detection parameter that defines the size of the face is set to 1/2 of the face size of the previous image, and the minimum value is set to the maximum value (for example, in the case of default) As described above, the range of the face size parameter close to the number of pixels on the short side of the image is narrowed. Further, a change is made such that the number of detected faces, which is the number of faces to be examined in the detection, is reduced to 10 from the number of detections set in the previous image (for example, 256 in the case of default).

  Alternatively, when the previously processed image is a group photo or the like, the detected face size is small, so the maximum value of the face size parameter is up to twice the face size parameter of the previous image. The range is limited, and the face direction is limited to the top-and-bottom direction only.

  Further, the detection parameter related to the detection area is controlled so that face detection is performed around the same area as the area where the face was detected in the previous image. For example, an area including an area where a face is detected in the previous image is set, and the detection area is limited so that face detection is performed for this area. Alternatively, it is possible to perform a masking process on a region other than a region including a region where a face has been detected in the previous image, and perform face detection with image data invalidated.

  As described above, when the correlation between the image to be processed and the previous image is high, the detection parameter range is narrowed. Thus, according to the parameters, when detecting a detection target such as a face by moving or changing a detection region such as a template, as described later, the range to be moved or changed can be narrowed, and the time required for detection processing can be reduced. It can be made shorter.

  After the detection parameter is changed as described above in step S107, or when it is determined in step S106 that there is no change in the detection parameter, face detection processing is performed using the detection parameter set at that time in step S108. This face detection is performed by using a detection region such as a template and moving or changing the detection region according to a detection parameter changed or not changed as described above. In addition, a well-known thing can be used for this process, The detailed description is abbreviate | omitted here.

  After the face detection process in step S108, the process proceeds to step S109 regardless of whether or not a face is detected there, and the shooting information and the detection result if a face is detected are stored in a predetermined memory, and the process proceeds to step S110. . In step S110, if a face is detected, image processing is performed on the face region image based on the result. This image processing includes so-called red-eye detection in the face area and correction processing for it, sharpening of the face area, conversely smoothing processing of the face area, etc., and known methods can be used. These corrections may be selected by the user.

(Second embodiment)
In the above-described first embodiment, the configuration using the imaging information and the face detection result of the image processed last time is shown. However, when there is almost no correlation between the previous image and the image to be processed, the face detection parameters used in the previous image are used as they are, and appropriate face detection can be performed for the image to be processed. Not likely to be. For this reason, in this embodiment, the detection parameter changing process in step S106 shown in FIG. 3 is performed using default detection parameters in accordance with the correlation.

  FIG. 7 is a flowchart showing detection parameter change processing according to the second embodiment of the present invention.

  As shown in the figure, two threshold values are provided for the correlation between images represented by the total evaluation amount described in FIG. 6, and the first threshold value is set to a value larger than the second threshold value. . In this case, the respective threshold values are set in two stages for the determinations 1 to 6 described above with reference to FIG.

  In step S403, it is determined whether the total evaluation amount is larger than the first threshold value. Here, when it is larger than the first threshold value, that is, when it is determined that the correlation between the images is extremely high, the process proceeds to step S404, and the detection parameter change flag is set to “1” (change).

  If the correlation is smaller than the first threshold value in step S403, the process proceeds to step 406. In step S406, it is determined whether or not the correlation is smaller than the second threshold value. Here, when it is smaller than the second threshold value, that is, when it is determined that there is almost no correlation between images, the process proceeds to step S407, and the detection parameter change flag is set to “2” (default). In other cases, the detection parameter change flag is set to “0” (not changed) in step S405.

  When the detection parameter change flag is set to “2” (set to default) by the above processing, the detection parameter can be returned to the default value in step S107 shown in FIG. That is, when there is almost no correlation between images, it is possible to perform detection using the default detection parameters without being based on the previous image, and when there is a moderate correlation (step S405), the correlation is Based on a previous image, it is possible to perform detection using detection parameters that are not changed.

(Other embodiments)
In the first and second embodiments described above, the detection parameters are controlled by paying attention to the correlation between images. However, various combinations of face detection parameters are prepared according to the photographing information. Alternatively, in step S106 in FIG. 3, it may be determined which combination of detection parameters is used to control the face detection parameters.

  In addition, as described with reference to FIG. 6, when using histogram analysis of image data for correlation evaluation, the amount of processing increases when histogram analysis is performed on all images. It may be created and the histogram analysis may be performed on the thumbnail image, in which case the analysis becomes more efficient. In addition to histogram analysis, frequency analysis, scene analysis, and the like can be employed as evaluation quantities.

(Still another embodiment)
The image processing apparatus described above can be applied to a system composed of a plurality of devices (for example, a host computer, an interface device, a reader, a printer, etc.), or a device (for example, a copier, a facsimile apparatus) composed of a single device. Etc.).

  In addition, an object of the present invention is to supply a storage medium (or recording medium) in which software program codes for realizing the functions of the embodiment described above with reference to FIG. Needless to say, this can also be achieved by a computer (or CPU or MPU) reading and executing the program code stored in the storage medium. In this case, the program code itself read from the storage medium realizes the functions of the above-described embodiments, and the storage medium storing the program code constitutes the present invention. Further, by executing the program code read by the computer, not only the functions of the above-described embodiments are realized, but also an operating system (OS) running on the computer based on the instruction of the program code. It goes without saying that a case where the function of the above-described embodiment is realized by performing part or all of the actual processing and the processing is included.

  Furthermore, after the program code read from the storage medium is written in the memory of the function expansion card inserted into the computer or the function expansion unit connected to the computer, the function expansion is performed based on the instruction of the program code. It goes without saying that the CPU or the like provided in the card or the function expansion unit performs part or all of the actual processing, and the functions of the above-described embodiments are realized by the processing.

1 is a block diagram illustrating a hardware configuration of a printing system according to an embodiment of the present invention. It is a figure which shows an example of the format 201 of the image data obtained by imaging | photography with the digital still camera 109 shown in FIG. It is a flowchart which shows the image process regarding the face detection accompanying the change process of the detection parameter which designates the area | region for a detection based on 1st embodiment of this invention. It is a flowchart which shows the imaging | photography information analysis process of step S104 shown in FIG. It is a flowchart which shows a judgment process whether the change of the face detection parameter of step S106 shown in FIG. 3 is required. In the process of step S302 in FIG. 5, evaluation amounts such as various shooting information such as a shooting mode and a shooting time, a histogram of each pixel value of image data, and each evaluation for determining the correlation between two adjacent images. It is a figure explaining the method of each judgment by the combination of quantity. It is a flowchart which shows the detection parameter change process which concerns on the 2nd Embodiment of this invention.

Explanation of symbols

101 Image processing apparatus 102 CPU
103 ROM
104 RAM
105 Secondary Storage Device 106 Display Device 107 Input Device 108 I / F Control Unit 109 Digital Still Camera 110 Memory Card Reader / Writer 111 Printer

Claims (15)

An image processing apparatus that detects a predetermined detection target in an image using a detection parameter that defines a detection processing range when detecting the detection target in the image,
Correlation determining means for determining the degree of correlation between an image on which the predetermined detection target is detected and an image on which the predetermined detection target is detected next to the image;
Correction means for correcting a detection parameter used for the detection in the next image for detecting the predetermined detection object in accordance with the degree of correlation determined by the correlation determination means;
An image processing apparatus comprising:   The correlation determination unit determines a correlation between the two images based on an evaluation amount determined for the image, and the correction unit corrects a detection parameter when the evaluation amount is larger than a predetermined threshold. The image processing apparatus according to claim 1.   The correlation determination unit determines a correlation between the two images based on an evaluation amount determined for the image, and the correction unit corrects a detection parameter when the evaluation amount is larger than a first threshold, and the evaluation amount The image processing apparatus according to claim 1, wherein a default detection parameter is set when is smaller than a second threshold smaller than the first threshold.   The image processing apparatus according to claim 1, wherein a photographing condition is used as the evaluation amount.   The image processing apparatus according to claim 1, wherein a photographing time is used as the evaluation amount.   The image processing apparatus according to claim 1, wherein a histogram analysis result of an image is used as the evaluation amount.   The image processing apparatus according to claim 6, wherein a reduced image of the image is created and the histogram analysis is performed. An image processing method for detecting a predetermined detection target in an image using a detection parameter that defines a detection processing range when detecting the detection target in the image,
A correlation determination step of determining a degree of correlation between an image on which the predetermined detection target is detected and an image on which the predetermined detection target is detected next to the image;
A correction step of correcting a detection parameter used for the detection in the image in which the predetermined detection target is detected next, according to the degree of correlation determined by the correlation determination unit;
An image processing method characterized by comprising:   The correlation determining step determines a correlation between the two images according to an evaluation amount determined for the image, and the correction step corrects a detection parameter when the evaluation amount is larger than a predetermined threshold value. The image processing method according to claim 8.   The correlation determination step determines a correlation between the two images based on an evaluation amount determined for the image, and the correction step corrects a detection parameter when the evaluation amount is larger than a first threshold, and the evaluation amount 9. The image processing method according to claim 8, wherein a default detection parameter is set when is smaller than a second threshold value smaller than the first threshold value.   The image processing method according to claim 8, wherein a photographing condition is used as the evaluation amount.   The image processing method according to claim 8, wherein a shooting time is used as the evaluation amount.   The image processing method according to claim 8, wherein a histogram analysis result of an image is used as the evaluation amount.   The image processing method according to claim 13, wherein a reduced image of the image is created and the histogram analysis is performed. A program for causing an information processing apparatus to execute image processing for detecting a predetermined detection target in an image using detection parameters,
A correlation determination step of determining a degree of correlation between an image in which the detection target is detected and an image in which the predetermined detection target is detected next to the image;
A correction step of correcting a detection parameter used for the detection in the image in which the predetermined detection target is next detected according to the degree of correlation determined by the correlation determination unit;
The program characterized by having.

Images