JP2012094985A - Image processing device, method, program, and storage medium - Google Patents

Abstract

PROBLEM TO BE SOLVED: To improve an image quality only in a required feature area to reduce a data quantity.SOLUTION: Moving image data is stored in a frame memory 14. A subject area detection part 27 detects a subject area including a subject from a frame image. A movement direction detection part 28 detects the direction of the movement of the subject on the basis of correlation between preceding and following frame images. A division part 29 divides a screen of a frame image into plural areas on the basis of the direction of the movement. A compressibility setting part 31 sets compressibility to each of the areas so that an area in the direction of the movement is improved in the image quality. A compression part 32 compresses each of the images of the areas in accordance with the compressibility.

Description

  The present invention relates to an image processing apparatus, method, program, and recording medium that compress an area obtained by dividing a screen into a plurality of areas by changing a compression rate.

  Conventionally, JPEG2000 (ISO / IEC FCD 15444-1) is known as an image compression method. In this compression method, a selective area image quality improvement function (ROI function) for reducing the compression area (image quality) of a region of interest (ROI) of an image without lowering the compression ratio of the entire image. There is. However, the attention area and the compression ratio of each area have to be set manually. Thus, an image processing apparatus that can be automatically set has been proposed (Patent Document 1).

  The image processing apparatus described in Patent Literature 1 detects a plurality of feature regions having different types of features, such as a moving object such as a person or a vehicle, from an image, and each detected feature region has a higher image quality than other regions. As described above, compression is performed with different compression strengths according to the types of features in the feature region.

JP 2009-49976 A

  However, in the invention described in Patent Document 1, since compression is performed with different compression strengths according to the features of the feature regions, the same type of feature regions are compressed with uniform compression strength, and even unnecessary regions are compressed. There is a risk of higher image quality (lower compression ratio). In this case, there is a disadvantage that the amount of data increases.

  The present invention has been made in view of the above problems, and an object of the present invention is to provide an image processing apparatus, method, program, and storage medium capable of improving the image quality of only necessary feature areas so that the amount of data is reduced. .

  In order to achieve the above goal, in the present invention, a moving direction detecting unit that detects a moving direction of a subject from an image; a dividing unit that divides the image into a plurality of regions according to the moving direction of the subject; Compression means for compressing each of the images at different compression ratios so that the region in the moving direction becomes a higher quality image.

  As the moving direction detection means, it is desirable to acquire at least information on the moving direction of the subject based on the correlation between the image and the image taken at a different time. Specifically, subject region detection means for detecting a region (subject region) including an image of the main subject is provided, and the subject region in a continuous frame of the current frame image constituting the moving image data and the previous previous frame image. The image signal level correlation is detected, and the motion vector of the main subject is obtained based on the detected correlation.

  As the dividing means, it is desirable to divide the subject region into a region in the moving direction of the subject (moving direction side region) and a region opposite to the moving direction. In this case, it is desirable that the compression means compresses the image of the region including the subject region on the moving direction side of the subject at a low compression rate so that the image is higher in quality than the region on the opposite side.

  The dividing means divides the subject area including the subject and the moving direction side area in the moving direction, and the compressing means compresses the images of the subject area and the moving direction side area by changing the compression rate. May be. In this case, set three levels of compression ratios, high, medium, and low, to set the subject area to a low compression ratio, the moving direction side area to a medium compression ratio, and the opposite direction side area to the highest image quality. What is necessary is just to set each to a high compression rate.

  By the way, the dividing means determines the number to be divided according to the position and size of the subject area and the moving direction. In the case of dividing three or more, a grouping unit that generates a group region by grouping the plurality of divided regions based on the moving direction may be provided. In this case, it is desirable that the compression unit compresses each of the group areas while changing the compression rate so that the image in the movement direction side area has high image quality.

  In addition, it is preferable that a compression rate setting unit that assigns a different compression rate based on the moving direction to each of the plurality of divided regions or the plurality of group regions is desirable.

  Furthermore, as a compression means, an image in a low compression rate area that is compressed at a low compression rate so as to achieve high image quality, and an image in a high compression rate area that is compressed at a high compression rate so as to obtain a lower image quality than that. A plurality of encoding means for performing parallel processing of encoding, a timing adjusting means for adjusting the transfer timing after the encoding process of each encoding means, and a combining means for combining a plurality of encoded data transferred from the timing adjusting means It is desirable to configure with

  In the present invention, the moving direction of the subject is detected from the image, the screen is divided into a plurality of screens based on the moving direction of the subject, and the images of the divided regions become higher in quality as the region is in the moving direction of the subject. As described above, since the compression is performed by changing the compression rate, only the image of the area required for the subject in the feature area can be improved in image quality, and thus the data amount can be reduced.

It is a block diagram which shows the structure of the digital camera which employ | adopted this invention. It is a block diagram which shows the structure of an image process part. It is a block diagram which shows the structure of a compression part. It is a flowchart figure which shows the procedure of the video recording operation | movement with a digital camera. It is explanatory drawing which shows the state which a to-be-photographed region detection part detects a to-be-photographed region by setting a rectangular frame. It is explanatory drawing which shows the state in which a moving direction detection part detects the moving direction of a to-be-photographed object based on the correlation of a front and back frame image, (a) is front image data, (b) is present image data. It is explanatory drawing which shows the state which a division part divides | segments a screen based on the moving direction of a to-be-photographed object. FIG. 10 is an explanatory diagram for explaining another example of selecting a main subject area from among a plurality of subject areas when detecting a plurality of subject areas. FIGS. 9A and 9B are explanatory diagrams illustrating a state in which the moving direction of the main subject area described with reference to FIG. 8 is detected based on the correlation between the previous and next frame images, where FIG. 9A is the previous image data and FIG. It is explanatory drawing which shows the state which divided | segmented the screen based on the moving direction demonstrated in FIG. It is explanatory drawing which shows the area | region grouped based on the division demonstrated in FIG. It is explanatory drawing which shows the example of a screen division | segmentation in the case of detecting a moving direction of a subject as an upward direction. It is explanatory drawing which shows the example of a screen division | segmentation in case a some main subject area | region is detected and the moving direction of each subject is detected as the opposite direction. It is explanatory drawing which shows the example which groups the division area demonstrated in FIG. FIG. 10 is an explanatory diagram illustrating another example in which priority is given to a movement direction side area when a movement direction side area and a non-movement direction side area of a subject overlap.

  As shown in FIG. 1, a digital camera 10 with a moving image recording and reproducing function that is an embodiment using the present invention includes an imaging unit 11, a control unit 12, a signal processing unit 13, a frame memory 14, and an image processing unit 15. ROM 16, RAM 17, I / F (interface) 18, VRAM 19, display image data generation unit 20, and display control unit 21, each of which is connected to a bus 22. The imaging unit 11 outputs digital image data obtained by imaging a subject and performing A / D conversion. The image data is a frame image or a field image. The control unit 12 controls each unit centrally. The operation unit 23 is connected to the control unit 12 and includes an operation button group for moving image recording, an operation button group for moving image reproduction, and the like.

  The signal processing unit 13 takes in digital image data and performs corrections such as pixel defect correction, white balance correction, and gamma correction. The frame memory 14 is constituted by a capacity semiconductor memory such as SDRAM, and stores image data for one frame or several frames corrected by the signal processing unit 13.

  The ROM 16 stores various programs and set values necessary for executing the programs in advance. The RAM 17 is used as a work memory for the control unit 12 and as a temporary memory for each unit. A recording unit 24 that records encoded image data is connected to the I / F 18. As the recording unit 24, a built-in semiconductor memory, a hard disk, or the like can be used. For external connection, the recording medium may be a removable recording medium, a slot for inserting the recording medium, a circuit for controlling access to the recording medium, and the like.

  The display image data generation unit 20 generates and outputs display image data by enlarging or reducing the image data for one frame read from the frame memory 14 based on the resolution and the number of pixels of the display unit 25. To do.

  The VRAM 19 is a display video memory that stores display image data. The display control unit 21 controls the display image data to be read from the VRAM 19 and displayed on the display unit 25. On the display unit 25, image data of a through image including image data being recorded, or image data of a moving image reproduced from the recording unit 24 is displayed.

  As shown in FIG. 2, the image processing unit 15 includes a subject area detection unit 27, a movement direction detection unit 28, a division unit 29, a grouping unit 30, a compression rate setting unit 31, a compression unit 32, and an expansion unit 33. ing.

  The subject region detection unit 27 detects a region where an image including a subject is captured from an image of one image data (current image data) as a subject region, and outputs image extraction of the subject region and position information thereof. Specifically, the subject region detection unit 27 detects a region including a human face, a human body, a pet, and a subject such as a vehicle or an airplane as a subject region.

  The moving direction detection unit 28, based on the current image data and the previous image data taken immediately before the current image data, for example, selects a matching region having the greatest correlation with the image of the subject region from the previous image data. By searching, the moving direction of the subject is detected.

  The dividing unit 29 divides an area other than the subject area in the screen of the current image data based on the moving direction of the subject. Specifically, the subject region is divided into a region on the moving direction side of the subject and a region on the opposite direction side.

  The grouping unit 30 groups a plurality of areas divided including the subject area based on the moving direction of the subject. Specifically, grouping is performed based on whether or not the target area belongs to the traveling direction side with respect to the subject area. For example, a region belonging to the traveling direction side is grouped into a traveling direction side group region, and a region belonging to the opposite direction side is grouped into a non-traveling direction side group region.

  The compression rate setting unit 31 sets the compression rate for each image in the group area based on information on the group type (whether or not the group is on the traveling direction side). For example, the compression rate is set so that the image in the group area in the traveling direction has low compression (high image quality), and the image in the group area in the non-travel direction side has high compression (low image quality) higher than the compression ratio in the traveling direction side area. Assign (image quality level).

  The compression unit 32 compresses the images in each group area in parallel at the assigned compression rate, synthesizes the compressed data obtained by compressing the image in the area, and outputs the combined compressed image data.

  The combined compressed image data is recorded in the recording unit 24 as combined compressed image data with a decoding table describing parameters to be referred to at the time of decoding as a header.

  As described above, in the present embodiment, encoding / decoding is performed with the grouped region images as one unit. The decompressing unit 33 decrypts the combined compressed data and outputs the decrypted image data.

  As shown in FIG. 3, the compression unit 32 includes an encoding unit 39, a timing adjustment unit 42, a synthesis unit 43, and a decoding table addition unit 44. The encoding unit 39 includes a plurality of encoders 40 and 41. The encoders 40 and 41 are provided in advance as many as the grouping unit 30 performs grouping, and perform compression processing in parallel for each image in the group area. Note that only one encoder may be provided to compress the images in each group region in time series.

  The timing adjustment unit 42 is configured by a buffer memory, takes in the compressed data output from each encoding, buffers it, and adjusts the timing for simultaneous output to the synthesis unit 43.

  The synthesizing unit 43 takes in a plurality of compressed data from the timing adjusting unit 42, synthesizes the compressed data, and outputs synthesized compressed data.

  The decoding table adding unit 44 generates a decoding table, adds the generated decoding table as a header to the combined compressed data, and outputs the combined compressed image data. Based on the information obtained from the grouping unit 30, the compression rate setting unit 31, and the encoding unit 39, the decoding table is based on the positional information, compression rate, resolution, bit depth of the moving direction side and non-moving direction side regions. , Color specifications, bit rate, and other prescribed parameters are assigned as a table. These parameters are used when the decompressing unit 33 performs decoding.

  Note that the encoders 40 and 41 sequentially perform DC level shift, inter-component transformation, discrete wavelet transformation, quantization, ebcot (entropy coding), and streaming, as is well known. In streaming, entropy-encoded data is streamed together with encoding parameters such as a quantization width, and compressed data is output. In addition, the encoders 40 and 41 integrate the code amount when streamed, and give the integrated code amount to the decoding table adding unit 44.

  Next, the operation of the above configuration will be described with reference to FIG. When a recording start button constituting the operation unit 23 is operated (S-1), the control unit 12 drives the imaging unit 11 to capture a subject image. Digital image data output from the imaging unit 11 is stored in the frame memory 14 (S-2).

  The image processing unit 15 captures image data for one frame stored in the frame memory 14. The subject area detection unit 27 detects the subject area based on the captured image of the current image data (S-3).

  For example, the image of the subject area is detected in the rectangular frame 50 with respect to the current image data 49 shown in FIG. 5, and the image in the rectangular frame 50 is extracted and stored in the RAM 17 as a reference image. Further, the position information of the rectangular frame 50 is also stored in the RAM 17. The position information is composed of the coordinate value of the pixel at the upper left corner of the rectangular frame 50 and the number of vertical and horizontal pixels in the area. Here, the example shown in FIG. 5 will be described assuming that only one subject area is detected.

  The moving direction detection unit 28 stores the previous previous image data 51 in the RAM 17, and sets the comparison area 52 having the same size as the subject area as the previous image data 51 with reference to the RAM 17 as shown in FIG. 6. The comparison area 52 is sequentially set to different coordinate positions in the previous image data 51, the matching area having the largest correlation with the pixel value of the subject area is specified, and the whole matching image of the specified matching area and the subject area is determined. The moving direction of the subject is detected based on the positional relationship (S-4), and the information is sent to the dividing unit 29. In the example shown in the figure, the moving direction of the subject is detected as “right direction”.

  The dividing unit 29 refers to the moving direction of the subject and the position information of the subject region, and as shown in FIG. 7, the dividing unit 29 moves rightward with respect to the X coordinate so that the subject region is included in the moving direction side region. The area 55 and the area 56 on the opposite side are divided by the boundary line 57 (S-5), and the position information of each of the divided areas 55 and 56 is sent to the grouping unit.

  The grouping unit 30 refers to the position information of the divided areas 55 and 56, and moves the area 55 on the moving direction side including the subject area (rectangular frame 50) into the traveling direction group area and the area 56 on the opposite direction side. Are grouped into non-traveling direction group regions (S-6). The group area type information is sent to the compression rate setting unit 31.

  Based on the group type, the compression rate setting unit 31 sets a low compression rate to the encoder 40 to which an image in the traveling direction group region is input and sets a high compression rate to the encoder 41 to which an image in the non-traveling direction group region is input. Each compression rate is set (S-7).

  The encoding unit 39 reads the current image data from the RAM 17, compresses the images in the traveling direction group area with the encoder 40 at a low compression rate (S-8), and the encoder 41 determines the non-traveling direction group area. The image is compressed at a high compression rate (S-9).

  A time difference occurs in the compression processing of the encoders 40 and 41. This time difference is absorbed by the timing adjustment unit 42. The timing adjustment unit 42 stores the compressed data from the plurality of encoders 40 and 41 until the output of the compressed data is complete, and sends the two compressed data to the synthesis unit 43 in response to the completion (S-10). The synthesizer 43 synthesizes the two compressed data to generate synthesized compressed data (S-11).

  The decoding table adding unit 44 generates a decoding table based on the position information of the group area obtained from the grouping unit 30 and the compression rate setting unit 31, the compression rate information for the position information, the bit rate information, and the like. The decoded table is added to the combined compressed data as a header and output as combined compressed image data (moving image encoded stream) (S-12). The output combined compressed image data is recorded in the recording unit 24. Thereafter, until the recording stop operation (S-13) is performed, the next frame image is compressed for each group area in the same manner as described above, and composite compressed image data is created.

  The combined compressed image data is read from the recording unit 24 in response to the reproduction operation, decoded by the decompression unit 33, and reproduced on the screen of the display unit 25.

  In the above-described embodiment, the image data to be compressed by the image processing unit 15 may be RAW data acquired by a three-plate image sensor having one solid-state image sensor for each RGB color, or may be a color filter array ( RAW data acquired by a single-plate image sensor having CFA) may be image data obtained by performing demosaicing processing and converting it to a full color image (RGB image).

  In addition, the subject region detection unit 27 of each of the above embodiments may detect a plurality of subject regions. In this case, it is preferable to select a main subject region (main subject region) from the region that is a moving body or the position and size of the subject region. In this case, selection means for comparing a plurality of subject areas and selecting the main subject area based on the position and size may be provided.

  As the selection means, for example, when the area is smaller than a predetermined reference size, or when it is close to the screen edge outside the predetermined central range with respect to the entire screen, it is selected as the main subject area. do not do.

  For example, in the current image data 60 shown in FIG. 8, the subject region detection unit 27 detects two subject regions 61 and 62. In this case, the selection means selects the subject area 61 as the main subject area based on the positions and sizes of the two subject areas 61 and 62. As shown in FIG. 9, the moving direction detecting means 28 detects the moving direction 63 of the subject image in the main subject region 61 based on the frame correlation with the previous image data 60a.

  As shown in FIG. 10, the dividing unit 29 sets boundary lines 64 and 65 orthogonal to or intersecting the detected moving direction 63 based on the subject areas 61 and 62, and displays the current image data 60 on a plurality of screens. Divide into 66-68. As shown in FIG. 11, the grouping unit 30 includes a region 66 including the main subject region 61 and a region 67 on the moving direction side in the traveling direction group region 69 and a region 68 on the reverse direction side in the non-traveling direction side. Each is set in the group area 70.

  In this case, since the image in the group area 69 is an area on the moving direction side including the main subject, it is stored with high image quality. On the contrary, the image of the group area 70 is not on the moving direction side of the main subject, and therefore any image appears here is stored with low image quality.

  When the moving direction detection unit 28 detects the moving direction of the subject as the upward direction, as shown in FIG. 12, the dividing unit 29 has a boundary line 72 orthogonal to or intersecting with the detected moving direction 71, that is, a horizontal line. A boundary line 72 in the horizontal direction is set based on the subject area 73 and divided into a plurality of screens 74 and 75. In this case, the grouping unit 30 sets the moving direction side area 74 including the subject area 73 as the traveling direction side group area.

  In the example shown in FIG. 13, since the two subject areas 80 and 81 are within the center range and exceed the reference size, the two subject areas 80 and 81 are respectively selected as the main subject areas. In the example shown in the figure, the movement directions 82 and 83 of the two main subjects are detected in opposite directions (left and right directions). In this case, the dividing unit 29 sets a boundary line between the two subject areas 80 and 81 and divides the screen 84 of the current image data into two areas. However, in this case, since both areas are on the moving direction side of the main subject areas 80 and 81, a low compression rate is set for both the left and right areas. For this reason, there is a possibility that the file size does not change compared to the case where the entire image is compressed at a low compression rate.

  Accordingly, the dividing unit 29 is configured to divide the entire image into at least three or more based on the sizes of the main subject areas 80 and 81 so that the file size becomes smaller than that obtained by compressing the entire image at a low compression rate. Is desirable. For example, as shown in FIG. 14, by setting regions 85 and 86 on the moving direction side in accordance with the vertical size of the main subject regions 80 and 81, the upper and lower portions of the main subject regions 80 and 81 are separated into different regions 87. ~ 90 can be set. Further, by dividing based on the horizontal size of the main subject areas 80 and 81, the area 91 between the main subject areas 80 and 81 can be set as another area. Therefore, in the example shown in the figure, the dividing unit 29 divides into nine regions 80, 81, 85-91.

  In this case, the grouping unit 30 may group the main subject region and the moving direction side region into the traveling direction side group region, and the other regions into the non-traveling direction side group region.

  In each of the above embodiments, the compression rate setting unit 31 has been described as an example in which the compression rate is set in two stages of a low compression rate and a high compression rate, but the present invention is not limited to this and is set in three or more steps. You can also For example, when the compression ratio is set to three levels of high, medium, and low, the grouping unit 30 has the two main subject areas 80 and 81 described in FIG. 14 as the low compression ratio group area on the moving direction side. Two areas 85 and 86 are set as medium compression ratio group areas, and the remaining areas 87 to 91 are set as high compression ratio group areas. In this manner, three or more compression rates may be set so that the area on the moving direction side of the subject has high image quality.

  In this case, of course, the encoding unit 39 is configured by three first to third encoders in advance according to the number of stages of the compression rate. Then, the compression rate setting unit 31 assigns a low compression rate to the first encoder that compresses the image in the low compression rate group region, and a medium compression rate to the second encoder that compresses the image in the medium compression rate group region. Then, a high compression rate is set for each of the third encoders that compress the images in the high compression rate group area.

  In each of the embodiments described above, the subject region detection unit 27 detects the subject region. However, the movement direction detection unit may detect both the subject region and its movement direction together. In this case, an area including a moving object is detected as a subject area based on the correlation between the previous and next frames of the moving image data. In the case of detecting a plurality of subject areas, for example, an area (moving area) in which the change amount of the pixel value is larger than a predetermined change amount between the current image data and the previous or subsequent image data is set as the main subject area. May be detected.

  Further, for example, in the case of moving image data obtained by photographing a landscape from a car window or moving image data obtained by panning a landscape, houses and trees in the foreground may be detected as the main subject area. In this case, since the main subject area continues in the horizontal direction of the screen and the subject area is not cut, the boundary to be divided may be determined with a predetermined length. That is, when the main subject area is detected to the full width of the screen, the horizontal or vertical length of the moving direction side area is, for example, a ratio of 2/3 to the horizontal or vertical length of the entire screen. What is necessary is just to predetermine the length etc. to become.

  Further, as shown in FIG. 15, for example, two subject areas 93 and 94 are detected at different positions on the screen, the moving direction of the subject in one area 93 is set to the right, and the moving direction of the subject in the other area 94 is set. Is detected as the vertical direction, the low compression region 94a on the moving direction side and the high compression region 93a on the non-moving direction side, which the dividing unit 29 divides based on the moving direction of the subject, may be considered. In this case, the grouping unit 30 may perform grouping by giving priority to the low compression rate area.

  As the subject region detection unit 27 in each of the above embodiments, a human body region including an object having a degree of matching with a pattern related to a person's body greater than a predetermined matching degree is detected by pattern matching or the like, and the detected region is used as a main subject. It may be detected as a region. Further, a face image detection unit may be provided instead of this detection unit. In this case, an area including a face image having a degree of coincidence with a pattern related to a predetermined person's face larger than the predetermined degree of coincidence is detected by pattern matching or the like, and the detected face image area is detected as a main subject area. do it. Note that the subject region detection unit may detect a human body region from a region existing in the vicinity of the face region.

  In addition, the subject region detection unit 27 extracts images (objects) included in each of a plurality of image data (frame images) by edge extraction or the like. Then, the subject area detection unit 27 identifies objects that are included in different positions of other image data and match with a degree of coincidence greater than a predetermined degree of coincidence, and an area including the identified object is identified. The main subject area may be specified.

  In addition to the above-mentioned person's face and human body, the subject area detection unit 27 also includes at least one part of the human head such as a part of the person's head or a part of the human body such as a person's hand or a living body other than the human body. An area where the part of the image is captured may be detected as the main subject area. Here, the living body includes a specific tissue existing inside the living body, such as a tumor tissue or a blood vessel inside the living body. Furthermore, the subject area detection unit may detect, as a main subject, an area in which a card such as money, a cash card, a vehicle, or a license plate of a vehicle is imaged in addition to a living body. Furthermore, the subject area detection unit may detect the main subject area based on the AF information.

  Further, the main subject area may be selected by the user designating a specific area on the current image. Furthermore, since the image of the main subject often appears in the center of the screen, a predetermined area such as the central area of the current image data may be automatically selected.

  In each of the above-described embodiments, the configuration is described in which moving image data is compressed by a digital camera. However, the present invention is not limited to this, and still image data can be compressed. In this case, the moving direction of the subject can be detected by designating and reading similar image data having the same shooting location and different shooting times, such as a plurality of image data shot continuously.

  In each of the above-described embodiments, the image data is compressed by the digital camera. However, the present invention is not limited to this, and the present invention can be applied to an apparatus configured by an application program executed by a personal computer or the like. In this case, an image file obtained by imaging with a digital camera or a video camera may be read by a personal computer or the like and compressed as described above.

  In the above embodiments, various embodiments and modifications have been described. However, the present invention is not limited to these contents. Other embodiments conceivable within the scope of the technical idea of the present invention are also included in the scope of the present invention.

DESCRIPTION OF SYMBOLS 10 Digital camera 15 Image processing part 27 Subject area detection part 28 Movement direction detection part 29 Dividing part 32 Compression part

Claims (12)

A moving direction detecting means for detecting the moving direction of the subject from the image;
Dividing means for dividing the image into a plurality of regions in accordance with the moving direction of the subject;
Compression means for compressing each of the images of the plurality of regions by changing the compression ratio so that the region in the moving direction of the subject becomes a high-quality image;
An image processing apparatus comprising: The image processing apparatus according to claim 1.
The image processing apparatus, wherein the moving direction detecting means detects a moving direction of the subject based on a correlation between the image and an image taken at a different time from the image. The image processing apparatus according to claim 1 or 2,
The dividing unit divides the moving direction side area in the moving direction of the subject and the area opposite to the moving direction;
The image processing apparatus according to claim 1, wherein the compression unit performs compression by changing a compression rate so that the moving direction side region has a higher quality image than the opposite side region. The image processing apparatus according to any one of claims 1 to 4,
An image processing apparatus comprising: a subject area detection unit that detects a subject area including an image of a subject from the image. The image processing apparatus according to claim 4.
The dividing means divides the subject area and a moving direction side area in the moving direction thereof,
The image processing apparatus, wherein the compression unit compresses the images in the respective regions at different compression ratios so that the subject region becomes a higher quality image than the moving direction side region. The image processing apparatus according to claim 4 or 5,
A grouping unit configured to group the plurality of divided areas based on a moving direction of the subject to generate a group area;
The image processing apparatus, wherein the compression unit compresses each of the images in the group area by changing a compression rate so that the moving direction area becomes a high-quality image. The image processing apparatus according to any one of claims 1 to 6,
An image processing apparatus comprising compression ratio setting means for assigning different compression ratios to the plurality of divided areas or the plurality of group areas including the subject area based on the moving direction of the subject. The image processing apparatus according to any one of claims 1 to 7,
The compression means includes
A plurality of encoding means for performing parallel processing on encoding of an image of a low compression rate setting area to be compressed so as to achieve high image quality and an image of a high compression rate setting area to be compressed to be lower in image quality;
Timing adjusting means for adjusting the transfer timing after the encoding process of each encoding means;
Combining means for combining a plurality of encoded data transferred from the timing adjusting means;
An image processing apparatus comprising: Detecting the moving direction of the subject from the image;
A dividing step of dividing the image into a plurality of regions based on the moving direction;
A compression step of compressing each of the images of the plurality of regions by changing a compression rate so that the region in the moving direction becomes a high-quality image;
An image processing method comprising: The image processing method according to claim 9.
The dividing step divides into a moving direction side region in the moving direction of the subject and a region opposite to the moving direction,
The image processing method characterized in that the compression step compresses each of the images in the region by changing the compression rate so that the moving direction side region becomes a higher quality image than the opposite side region. .   A computer-readable program for executing the image processing method according to claim 9 or 10.   The recording medium which recorded the computer-readable program of Claim 11.

Images