JP2009260633A - Image compressor, photographing and recording device provided with image compressor, and image compression program - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To solve problems that image compression technologies using human visual characteristics related to luminance are not available when taken images are compressed without interpolation and extrapolation. <P>SOLUTION: An image compressor has an orthorhombic lattice resampling part 2 which resamples imaging signals to lattice points of an orthorhombic lattice to generate orthorhombic lattice pixels that have luminance signals and color signals, an othorhombic lattice image processor 3 which performs image processing for the orthorhombic lattice pixels to generate orthorhombic lattice image signals, and an orhorhombic lattice image data compression part 4 which image compresses the orhorhombic lattice image signals to decrease contrast resolution of high frequency components of at least luminance signals and generates orthorhombic lattice image compression data. <P>COPYRIGHT: (C)2010,JPO&INPIT

Description

  The present invention relates to an image compression apparatus and an image compression program. More specifically, the present invention relates to processing for resampling a captured image to a lattice point of an orthorhombic lattice, and includes such an image compression apparatus. The present invention relates to an imaging recording apparatus.

  An apparatus capable of compressing and recording a captured image and reproducing it is known (see Patent Document 1). In the imaging compression recording / reproducing apparatus described in Patent Document 1, an imaging unit in which at least one of a plurality of imaging elements is arranged at a position shifted in phase from a position corresponding to another imaging element, and an image from the imaging unit An image recording / reproducing unit for recording compressed image data from the recording unit, reading compressed image data from the recording medium, and decompressing the compressed image data from the image recording / reproducing unit A decompressing unit; and an interpolating and synthesizing unit that performs interpolation and synthesis processing on the image from the image decompressing unit to increase the resolution.

  In recent years, the image sensor (imaging means) can be obtained by subjecting an image signal output from the image sensor and / or image processing to the image signal by using a multi-pixel method or a pixel shifting method using a three-plate method. The data rate of the image signal is increasing.

  In addition, with the increase in the data rate of these image pickup signals and / or image signals, the data rate should be increased so that the recording unit also conforms to this, but the video standard or technical background, In the present situation where the data rate of the recording unit is not sufficiently increased due to the cost background, the compression unit needs to compress the data rate to a desired data rate.

  However, the data rate compression (irreversible compression) by the compression unit will cause a considerable deterioration in image quality, and even if the above-mentioned increase in image resolution and improvement in the characteristics of the luminance frequency modulation degree due to the increase in the number of pixels are achieved. If there is image quality degradation due to this data rate compression, it will be a trade-off between higher resolution of the image quality and various image quality degradation due to compression, so the total image quality will not be improved.

  Therefore, in the technique described in Patent Document 1, an image output from an image sensor is sent to a compression recording unit without being interpolated and compressed in an imaging unit to which pixel shift is applied, and is recorded on a recording medium. When the recorded compressed image data is reproduced, pixel interpolation is performed after the compressed image data is expanded in the reproduction unit. That is, by compressing and recording a captured image obtained by pixel shifting without performing interpolation and synthesis, the captured image is not up-converted to increase the data rate.

  As described above, according to the imaging compression recording / reproducing apparatus described in Patent Document 1, the input data rate to the compression recording unit is reduced, and the compression rate when compressing to a desired data rate can be reduced.

Here, elements related to a general image compression technique are mainly as shown in the following (1) to (4).
(1) Image compression on the planar phase when there is no change in adjacent pixels (2) Image compression on the time axis when there is no change between frames (3) Compactly expressing data patterns that are repeated statistically frequently (4) Image compression using human visual characteristics
JP 2007-67823 A

  However, as in the imaging compression recording / reproducing apparatus described in Patent Document 1, when image compression is performed without performing interpolation synthesis on a captured image obtained by pixel shifting, among the image compression (1) to (4) described above, For (1) to (3), a certain degree of improvement in compression efficiency can be aimed at, but for (4), the compression efficiency cannot be expected to be improved so much.

  This is because image compression without interpolating and synthesizing a captured image obtained by pixel shifting means that image compression is performed before the luminance signal component is generated. This is because compression cannot be carried out actively. For example, if image compression is performed before the luminance signal component is generated, image compression obtained by degrading the contrast resolution of the luminance high frequency component cannot be performed.

  Furthermore, if the image is compressed before the luminance signal component is generated, various restrictions are imposed on the specific image processing relating to the luminance. For example, specific image processing relating to luminance includes color reproducibility processing relating to KNEE processing in a high luminance region, desaturation processing of color false components generated in high frequency components of luminance, and the like. If the image is postponed later, aggressive image compression cannot be applied to the portion to be rotated later. Alternatively, if the image compression is to be positively performed, various color signals and / or color difference signals associated with the luminance due to deterioration in calculation accuracy (decrease in the number of calculation bits) due to image compression, There should be concern about significant image quality degradation.

  As described above, the imaging compression recording / reproducing apparatus according to Patent Document 1 can prevent an increase in the input data rate of the captured image to the compression recording unit by performing pixel interpolation synthesis in the reproduction unit. There is a problem that it is impossible to use an image compression technique that positively utilizes human visual characteristics.

  Accordingly, the present invention has been made in view of the above-described problems, and reduces the input data rate when compressing a captured image to reduce the image compression rate, and also improves human visual characteristics related to luminance. To provide an image compression apparatus using an image compression technique with higher compression efficiency and an image recording apparatus equipped with such an image compression apparatus by using the image compression technique actively used in combination. Objective.

  An image compression apparatus according to the present invention includes an orthorhombic grid resampling unit that resamples an imaging signal to a grid point of an orthorhombic grid to generate an orthorhombic grid pixel having a luminance signal and a color signal; Image processing is performed to generate an orthorhombic lattice image signal, and image compression that lowers the contrast resolution of at least the high-frequency component of the luminance signal is applied to the orthorhombic lattice image signal. And a rhombic lattice image data compression means for generating rhombic lattice image compression data.

  The imaging recording apparatus according to the present invention includes the above-described image compression apparatus, imaging means for photoelectrically converting an optical image for each imaging element to generate an imaging signal, and an oblique lattice for recording the oblique lattice image compression data. And an image recording means.

  Further, the image compression program according to the present invention includes a step of resampling an imaging signal to a grid point of an orthorhombic lattice to generate an orthorhombic lattice pixel having a luminance signal and a color signal; And generating an orthorhombic lattice image signal and applying image compression to the orthorhombic lattice image signal to lower the contrast resolution of at least the high frequency component of the luminance signal. Generating the step of causing the computer to execute.

  According to the image compression apparatus, the imaging recording apparatus, and the image compression program according to the present invention, the input data rate when compressing the captured image is reduced, and the image compression that positively uses the human visual characteristic regarding luminance is performed. By using, compression efficiency can be improved.

  Embodiments of the present invention will be described below with reference to the drawings.

-First embodiment-
FIG. 1 is a block diagram showing a configuration of an image recording apparatus according to the first embodiment of the present invention. An imaging recording apparatus according to the first embodiment includes an image sensor (imaging unit) 1, an orthorhombic lattice resampling calculation unit (orthogonal lattice resampling unit) 2, and an orthorhombic lattice image processing unit (orthogonal lattice image). A processing unit 3, an orthorhombic lattice image data compressing unit (orthogonal lattice image data compressing unit) 4, and an orthorhombic lattice image recording unit (orthogonal lattice image recording unit) 5. Among these, the rhombic lattice resampling calculation unit 2, the rhombic lattice image processing unit 3, and the rhombic lattice image data compression unit 4 constitute an image compression apparatus.

  The image sensor 1 photoelectrically converts an optical image for each imaging element to generate an imaging signal. The orthorhombic grid resampling calculation unit 2 resamples the imaging signal to grid points of the orthorhombic grid to generate an orthorhombic grid pixel having a luminance signal and a color signal.

  The orthorhombic lattice image processing unit 3 performs image processing on the orthorhombic lattice pixels to generate an orthorhombic lattice image signal. Here, if the re-sampling number of grid points is smaller than the number of elements of the image sensor, the substantial input / output data rate of the rhombic image processing unit 3 can be suppressed.

  The orthorhombic lattice image data compression unit 4 performs image compression on the orthorhombic lattice image signal to generate orthorhombic lattice image compression data. In the scanning direction of this image compression, Many lattices are included. In addition, the rhombic lattice image data compression unit 4 actively uses human visual characteristics related to luminance, and uses other image compression techniques for reducing the contrast resolution of luminance high frequency components of the rhombic lattice image signal. Used in combination with compression technology. The rhombic lattice image recording unit 5 records the rhombic lattice image compressed data.

  As described above, in the imaging / recording device according to the first embodiment of the present invention, after re-sampling the imaging signal output from the image sensor 1 to the grid points of the orthorhombic lattice, The image is recorded after being subjected to image compression including many rhombic lattice scans.

  FIG. 2A is a diagram partially showing a general Bayer array of the image sensor 1. In FIG. 2A, this Bayer array is a Bayer array composed of a square lattice, and includes a green filter (G), a red filter (R), and a blue filter (B).

  In FIG. 2A, a Bayer array composed of a square lattice is shown as a specific example. However, in the image sensor 1 included in the imaging recording apparatus according to the first embodiment of the present invention, a square lattice is used. In addition, the image sensor is not limited to the Bayer array, and an image sensor including an orthorhombic lattice and / or an array other than the Bayer array may be used.

  FIG. 2B is a diagram illustrating an example in which the lattice points of the rhombic lattice in the first embodiment of the present invention are the same as the array of the green filters of the Bayer array illustrated in FIG. It is.

  In FIG. 2B, the positions of the lattice points of the rhombic lattice indicated by the circles are the positions of the image sensor on which the green filter (G) is arranged, and the red filter (R) and the blue filter (B) The position of the image sensor on which the) is arranged is a position that is at the intersection of the horizontal auxiliary line indicated by the broken line L1 and the vertical auxiliary line indicated by the broken line L2, and is not a lattice point of the rhombic lattice. The grid points of the orthorhombic lattice are an example of the grid point array of the orthorhombic lattice used by the orthorhombic lattice resampling calculation unit 2, the orthorhombic lattice image processing unit 3, and the orthorhombic lattice image data compression unit 4, respectively. Is shown as

  FIG. 3A is a diagram illustrating an example of resampling each color imaging signal with respect to the lattice points of the orthorhombic lattice illustrated in FIG. Resampling to the lattice points of the orthorhombic lattice is performed by the orthorhombic lattice resampling calculation unit 2.

  In FIG. 3A, the resampling calculation to the lattice points of the rhombic lattice is performed by a plurality of photoelectric elements such as an image sensor (for example, a photodiode) in which a red filter (R) and a blue filter (B) are arranged. The imaging signal obtained by the conversion element) is subjected to planar phase conversion (positioning) to a lattice point array in which the green filter (G) is arranged.

  In FIG. 3A, the state after this plane phase conversion is shown as (G, R ′, B ′). This (R ′, B ′) indicates that each pixel has undergone pixel interpolation. As a specific method of this plane phase conversion, for example, a known method such as bicubic pixel interpolation can be used.

  FIG. 3B is a diagram illustrating an example in which the luminance signal Y and the color signal C are generated for the lattice points of the orthorhombic lattice illustrated in FIG. As shown in FIG. 3B, the orthorhombic lattice resampling calculation unit 2 performs the planar phase conversion (G, R ′, B ′) of each color imaging signal as shown in FIG. Based on this (G, R ′, B ′), a matrix operation of the luminance signal Y and the color signal C is performed to generate an orthorhombic lattice pixel represented by (Y, C). The color signal C may be replaced with color difference signals (Cr, Cb).

  The orthorhombic re-sampling operation unit 2 includes an arithmetic unit that performs plane phase conversion of each color imaging signal, and an arithmetic unit that generates a luminance signal Y and a color signal C from each color imaging signal (for example, a luminance matrix arithmetic unit and chroma). It is also possible to employ a configuration in which a matrix computing unit is provided as an integrated composite computing unit. If the configuration includes an integrated composite arithmetic unit that simultaneously calculates the plane phase conversion and the luminance signal Y and the color signal C, the entire imaging and recording apparatus of the present invention can reduce costs and downsizing. It is also possible to contribute.

  The imaging / recording apparatus of the present invention is limited to a configuration in which an arithmetic unit that performs phase phase conversion of each color imaging signal and an arithmetic unit that generates the luminance signal Y and the color signal C from each color imaging signal are integrated. Is not to be done.

  In addition, since the rhombic lattice image processing unit 3 and the rhombic lattice image data compression unit 4 perform processing based on the luminance signal Y, the image compression technique that positively utilizes human visual characteristics related to luminance is also used. When used, the calculation for generating the luminance signal Y from each color imaging signal is an essential calculation.

  If attention is paid to the horizontal auxiliary line (broken line L1) and the vertical auxiliary line (broken line L2) shown in FIG. 3B, the horizontal auxiliary line (dashed line 1) and vertical auxiliary line (broken line) shown in FIG. Compared with L2), it can be seen that there is no increase or decrease in the number of each.

  The fact that there is no increase / decrease in the horizontal auxiliary line (broken line L1) and the vertical auxiliary line (broken line L2) means that the rhomboid re-sampling calculation unit 2 in the first embodiment has taken images (imaging signals) of the same subject. ), The number of pixels (data amount) used in the orthorhombic lattice image processing unit 3 is substantially halved without substantially degrading the horizontal resolution and vertical resolution of the captured image, and the orthorhombic lattice image processing unit The data rate of the picked-up image input to 3 is substantially halved.

  FIG. 4A is a diagram illustrating an example of a zigzag scan (compression scan path) for general image compression processing on a conventional square lattice image. In FIG. 4A, this square lattice image has 8 × 8 pixels as a unit block image, and the unit block image composed of 8 × 8 pixels is zigzag scanned like a single stroke. Thus, a general zigzag scan often scans a square lattice in an oblique direction that is not the lattice.

  FIG. 4B is a diagram illustrating an example of a zigzag scan when compressing an orthorhombic lattice image according to the first embodiment of the present invention. The image compression process is performed by the rhombic lattice image data compression unit 4.

  In FIG. 4B, this orthorhombic lattice image has 2 × 4 × 4 pixels in which two sets of 4 × 4 pixels exist as a unit block, and the 2 × 4 × 4 pixels are written as one stroke. Zigzag scanning. As is apparent from FIG. 4B, the zigzag scan in the first embodiment includes more scans along the rhombic lattice than scans not along the lattice. By doing so, the luminance frequency modulation degree (or reproducibility) in the horizontal direction and the vertical direction of the image is improved as compared with the case of scanning along a square lattice.

  Here, when the zigzag scan of the conventional square lattice image shown in FIG. 4A is compared with the zigzag scan of the oblique lattice image according to the first embodiment of the present invention shown in FIG. The number of pixels in the lattice image is twice as large as the number of pixels in the oblique lattice image, and the number of scans (number of scans) is about twice as many.

  FIG. 5A and FIG. 5B are diagrams illustrating an example of a relationship between a zigzag scan for image compression processing and the position of the image sensor on which the green filter (G) is arranged. The zigzag scan shown in FIG. 5 (a) is equivalent to the conventional zigzag scan shown in FIG. 4 (a), and the zigzag scan shown in FIG. 5 (b) is the first of the present invention shown in FIG. 4 (b). This is equivalent to the zigzag scan in the embodiment.

  In FIGS. 5A and 5B, G indicated by a circle indicates the arrangement of actual pixels by the image sensor on which the green filter (G) is disposed, and G indicated by a broken circle. 'Indicates an array of interpolated pixels by pixel interpolation in which the green filter (G) is not mounted. A scan (scan) indicated by a broken line arrow indicates a scan between the G ′ interpolation pixels.

  Here, the number of scans by the conventional zigzag scan shown in FIG. 5A is compared with the number of scans by the zigzag scan in the first embodiment of the present invention shown in FIG. 2 times. Accordingly, with respect to the time lag (compression processing time) that occurs during compression scanning, the processing time of FIG. 5A is about twice that of FIG. 5B.

  However, since the actual luminance sampling number obtained from the image sensor 1 has a high correlation with the number of actual pixels by the image sensor on which the green filter (G) is mounted, as shown in FIG. Even if scanning is performed between G ′ interpolated pixels generated by pixel interpolation in which the green filter (G) is not mounted, improvement in luminance frequency modulation degree in the horizontal and vertical directions of this image signal can hardly be expected.

  Here, G ′ indicated by a broken line circle shown in FIG. 5A and a scan between G ′ interpolation pixels as indicated by a broken line arrow are excluded from FIG. 5A and indicated by a circle. If attention is paid only to the scan indicated by G and a solid arrow, it is understood that FIGS. 5A and 5B are geometrically similar. That is, if the essential luminance frequency characteristics are considered, it is shown that FIG. 5A and FIG. 5B can be considered almost equivalent.

  That is, the rhombic lattice image data compression unit 4 according to the first embodiment of the present invention hardly deteriorates the luminance frequency modulation degree in the horizontal direction and the vertical direction of the captured image in the captured image of the same subject. The time lag (compression processing time) generated during compression is substantially halved. The advantages obtained by reducing the time lag will be described later in the second embodiment of the present invention.

  The orthorhombic lattice image data compression unit 4 converts the orthorhombic lattice image signal into an integer lattice (including fixed-point conversion), and then uses the high-speed arithmetic circuit configured by addition / subtraction and bit shift to compress the oblique lattice image data. Is generated. Thereby, the further shortening effect of the time lag which generate | occur | produces at the time of compression is realizable. However, the rhombic lattice image data compression unit 4 is not limited to such a configuration.

  FIGS. 6A and 6B are examples of zigzag scanning when 2 × 2 × 2 pixels are used as unit blocks (shown as four unit blocks) in the first embodiment of the present invention. FIG. As shown in FIGS. 6A and 6B, the rhombic lattice image does not limit 2 × 4 × 4 pixels as a unit block. For example, any 2 × H × V pixels (where H and V are natural numbers, H is the number of pixels on the horizontal auxiliary line, and V is V It may be changed to a unit block consisting of the number of pixels on the vertical auxiliary line).

  As described above, even in the zigzag scan using 2 × 2 × 2 pixels as a unit block, the number of scans along the rhombic lattice is larger than the scan not along the lattice. The luminance frequency modulation characteristic (or reproducibility) in the vertical direction can be improved. In addition, the time lag that occurs during compression can be suppressed without substantially degrading the luminance frequency modulation degree in the horizontal and vertical directions.

  As described above, according to the image compression apparatus and the image recording apparatus including the image compression apparatus according to the first embodiment of the present invention, the luminance signal is obtained by resampling the imaging signal to the lattice points of the rhombic lattice. And an orthorhombic re-sampling operation unit 2 that generates an orthorhombic lattice pixel having color signals, an orthorhombic lattice image processing unit 3 that performs image processing on the orthorhombic lattice pixel to generate an orthorhombic lattice image signal, An orthorhombic lattice image data compression unit 4 is provided that generates orthorhombic lattice image compression data by applying image compression that lowers the contrast resolution of at least the high-frequency component of the luminance signal to the orthorhombic lattice image signal. Thereby, the input data rate at the time of compressing and recording a captured image can be reduced, the image compression rate can be reduced, and the data amount of image recording can be reduced. In addition, a compression technique with higher compression efficiency can be provided by using in combination with a compression technique that positively utilizes human visual characteristics related to luminance.

  In the above description, hardware processing is assumed as processing performed by the image compression apparatus. However, the processing is not limited to such a configuration. For example, a configuration in which processing is performed separately by software is also possible. In this case, the image compression apparatus includes a main storage device such as a CPU and a RAM, and a computer-readable storage medium in which a program for realizing all or part of the above processing is stored. Here, this program is called an image compression program. The CPU reads out the image compression program stored in the storage medium and executes information processing / calculation processing, thereby realizing the same processing as that of the above-described image compression apparatus.

  Here, the computer-readable recording medium refers to a magnetic disk, a magneto-optical disk, a CD-ROM, a DVD-ROM, a semiconductor memory, and the like. Alternatively, the image compression program may be distributed to a computer via a communication line, and the computer that has received the distribution may execute the image compression program.

  Hereinafter, a processing procedure of image compression processing realized by the CPU executing the image compression program will be described with reference to FIG.

  First, in step S1 of FIG. 7, the imaging signal is resampled to the lattice points of the rhombic lattice to generate an oblique lattice pixel having a luminance signal and a color signal. In step S2, image processing is performed on the orthorhombic lattice pixels to generate an orthorhombic lattice image signal. In step S3, the rhombic lattice image signal is subjected to image compression that lowers the contrast resolution of at least the high-frequency component of the luminance signal, thereby generating rhombic lattice image compression data, and this processing ends.

-Second Embodiment-
Next, a second embodiment of the present invention will be described. In the second embodiment, the rhombic lattice image data obtained by resampling to the lattice points of the orthorhombic lattice is temporarily recorded, and then the orthorhombic lattice image data is reproduced and re-applied to the lattice points of the square lattice. Resampling is performed to reproduce and output a square lattice image (for example, a monitor output image). In the following, in order to avoid duplication of explanation, the same components as those of the imaging and recording apparatus in the first embodiment are denoted by the same reference numerals, and detailed description thereof is omitted.

  FIG. 8 is a block diagram showing the configuration of the imaging / recording apparatus according to the second embodiment of the present invention. The imaging / recording apparatus according to the second embodiment has a configuration in which a reproduction block group is added to the configuration of the imaging / recording apparatus according to the first embodiment shown in FIG. The reproduction block group includes an orthorhombic lattice image data decoder (orthogonal lattice image data decoding unit) 11, a square lattice resampling unit (square lattice resampling unit) 12, and a square lattice image processing unit (square lattice image). Processing means) 13. The rhombic lattice image data decoder 11, the square lattice resampling calculation unit 12, and the square lattice image processing unit 13 are also components of the image compression apparatus.

  The orthorhombic lattice image data decoder 11 decodes the orthorhombic lattice image compressed data sequentially input from the orthorhombic lattice image data compressing unit 4 or the orthorhombic lattice image recording unit 5, and converts the orthorhombic lattice image data into Generate. The decoding of the rhombic lattice image compressed data, that is, the process of generating the rhombic lattice image data before compression from the rhombic lattice image compressed data can be performed using a known method.

  The square grid resampling unit 12 resamples the square grid image data to grid points of the square grid to generate square grid pixels. The square lattice image processing unit 13 performs image processing (square lattice image processing) on the square lattice pixels to generate a square lattice image (monitor output image).

  The number of square lattice pixels generated by the square lattice resampling calculation unit 12 can be substantially equal to the number of elements of the image sensor 1 (imaging device). This is because, as already described with reference to FIGS. 2B, 3B, 5A, and 5B, the image signal is represented by approximately half the number of lattice points of the image sensor 1. This is because even if re-sampling is performed, the horizontal resolution and vertical resolution of the captured image of the same subject do not deteriorate.

  In addition, as described in the first embodiment, the rhombic lattice image data compression unit 4 is resampled to approximately half the lattice points of the number of elements of the image sensor 1 described above. Is approximately halved. By adopting such a specification, it is possible to improve the real-time property of an electronic viewfinder or the like that outputs a monitor output image during imaging.

  As shown in FIG. 8, the monitor output image output from the electronic viewfinder during imaging is not transmitted from the rhombic lattice image recording unit 5 and is output from the rhomboid image data compression unit 4. It is generated via a route directly connected to the decoder 11.

  Here, the difference between the rhombic lattice image processing unit 3 and the square lattice image processing unit 13 will be described. As shown in FIG. 3B, the luminance signal Y is generated for each lattice point of the rhombic lattice. At this time, with respect to the luminance gradation adjustment of the luminance signal Y, it is preferable to perform the desired black adjustment and the high luminance KNEE process in the rhombic lattice image processing unit 3. This is because, in the final output image (for example, the monitor output image), an unnecessary signal component (a signal component less than the black offset or a resolution degradation component of a high luminance signal such as the KNEE region) is an oblique lattice image signal. This is because the dynamic range of the rhombic lattice image data compression unit 4 and the rhombic lattice image recording unit 5 can be effectively utilized.

  On the other hand, it is preferable that the correction of the modulation degree characteristic with respect to the luminance frequency and / or the correction of the contour characteristic is performed by the square lattice image processing unit 13. This is because when the modulation degree characteristic and the contour characteristic are corrected by the orthorhombic lattice image processing unit 3, the correction signal itself is resampled to the lattice points of the square lattice. This is because it is not a correct correction signal. One reason for not being an ideal correction signal is that when a correction signal consisting of one pixel on a lattice point of an orthorhombic lattice is resampled to a lattice point of a square lattice, the frequency of the correction signal is lowered. This is because the correction signal is a coarse correction signal composed of at least two pixels.

  FIG. 9A is a diagram partially showing an example of input data (orthogonal lattice image data) of the square lattice resampling calculation unit 12 in the second embodiment of the present invention. FIG. 9B is a diagram partially showing an example of output data (square lattice pixels) of the square lattice resampling calculation unit 12 in the second embodiment of the present invention.

  In FIG. 9A, the rhombic lattice image data (Y, C) is arranged on lattice points of a square lattice formed by the intersections of the horizontal auxiliary line indicated by the broken line L1 and the vertical auxiliary line indicated by the broken line L2. Is done. Then, as shown in FIG. 9B, interpolation pixels (Y ′, C ′) obtained by interpolation from the rhombic lattice image data are present at the lattice points of the square lattice where the rhombic lattice image data is not arranged. By being generated and arranged, image data (including interpolation pixels) is supplemented to all grid points of the square grid.

  As described above, according to the image compression apparatus and the image recording apparatus including the image compression apparatus in the second embodiment of the present invention, the rhombic lattice image data generated by the rhomboid image data compression unit 4 is compressed. An orthorhombic lattice image data decoder 11 that decodes the data to generate orthorhombic lattice image data, and a square lattice resampler that resamples the orthorhombic lattice image data to lattice points of a square lattice to generate square lattice pixels. In addition to the effects described in the first embodiment, the sampling calculation unit 12 and the square lattice image processing unit 13 that performs square lattice image processing on the square lattice pixels to generate a square lattice image are further provided. A square lattice image generation device can also be provided.

-Third embodiment-
Next, a third embodiment of the present invention will be described. The image capturing / recording apparatus according to the third embodiment compresses the image by further providing a feedback circuit between the structure for performing image capturing / recording and the structure for performing playback with respect to the image capturing / recording apparatus according to the second embodiment. The compression efficiency is further improved. Below, in order to avoid duplication of description, the same components as those of the imaging and recording apparatus according to the second embodiment are denoted by the same reference numerals, and detailed description thereof is omitted.

  FIG. 10 is a block diagram showing the configuration of the imaging / recording apparatus according to the third embodiment of the present invention. In the imaging / recording apparatus according to the third embodiment, the rhombic lattice image data compression unit 4 included in the imaging / recording apparatus according to the second embodiment illustrated in FIG. 8 includes a quantization table 40. The quantization table 40 is a data conversion table used when generating the rhombic image compression data. Here, the code of the rhombic lattice image data compression unit provided with the quantization table 40 is represented as 4A.

  The quantization table 40 defines 0 of the rhombic image compression data, that is, zero which is a reference value of the compressed data. The square lattice image processing unit 13 feeds back noise information of the square lattice image to the quantization table 40 (feedback). The rhombic lattice image data compression unit 4A can change the definition of 0 of the rhombic lattice image compression data generated by performing the compression process using the quantization table 40 based on the noise information. is there. With this feedback circuit, it is possible to prevent a reduction in the compression efficiency of image compression due to low luminance noise.

  This noise information is, for example, information obtained by quantifying the amount of dark current noise that stands out in the low luminance region of the luminance signal Y. This dark current quantification information is, for example, information obtained by measuring the frame correlation over several seconds in the moving image recording and recording apparatus. In the still image recording apparatus, for example, information obtained by capturing an all black image immediately before and / or immediately after a captured image to be recorded and measuring an amplitude width in one frame of the all black image. is there.

  As described above, in the third embodiment, the amount of noise in the low luminance region of the luminance signal Y is fed back to the quantization table 40 as information, and the rhombic lattice image generated by using the quantization table 40 is compressed. The definition of data 0 can be optimized, and the image compression efficiency can be further increased.

  As described above, according to the image compression apparatus and the image recording apparatus including the image compression apparatus according to the third embodiment of the present invention, the rhombic lattice image data compression unit 4 stores the rhombic lattice image compression data. The square lattice image processing unit 13 includes a quantization table 40 that defines a reference value 0, the noise information of the square lattice image is fed back to the quantization table 40, and the oblique lattice image data compression unit 4 is based on the noise information. Since the definition of the reference value 0 of the quantization table 40 is changed, the image compression efficiency can be further increased.

  The present invention is not limited to the above-described embodiments, and various modifications and applications can be made without departing from the gist of the present invention.

It is a block diagram which shows the structure of the imaging recording device in the 1st implementation of this invention. FIG. 2A is a diagram partially showing a general Bayer array of the image sensor. FIG. 2B is a diagram illustrating an example in which the lattice points of the orthorhombic lattice in the first embodiment are the same as the array of the Bayer array green filter illustrated in FIG. FIG. 3A is a diagram illustrating an example in which each color imaging signal is resampled with respect to the lattice points of the orthorhombic lattice illustrated in FIG. FIG. 3B is a diagram illustrating an example in which the luminance signal Y and the color signal C are generated for the lattice points of the orthorhombic lattice illustrated in FIG. FIG. 4A is a diagram illustrating an example of a general image compression zigzag scan for a square lattice image. FIG. 4B is a diagram illustrating an example of a zigzag scan when the rhombic lattice image is compressed in the first embodiment. FIG. 5A and FIG. 5B are diagrams illustrating an example of a relationship between a zigzag scan for image compression and the position of the image sensor on which the green filter is arranged. FIG. 6A and FIG. 6B are diagrams illustrating an example of a zigzag scan using 2 × 2 × 2 pixels as a unit block in the first embodiment. The flowchart which shows the process sequence of the image compression process implement | achieved by running an image compression program It is a block diagram which shows the structure of the imaging recording device in the 2nd implementation of this invention. FIG. 9A is a diagram partially illustrating an example of input data of the square lattice resampling calculation unit in the second embodiment. FIG. 9B is a diagram partially illustrating an example of output data of the square lattice resampling calculation unit according to the second embodiment. It is a block diagram which shows the structure of the imaging recording device in the 3rd implementation of this invention.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 ... Image sensor 2 ... Orthogonal lattice resampling calculating part 3 ... Orthogonal lattice image processing part 4 ... Orthogonal lattice image data compression part 5 ... Orthogonal lattice image recording part 11 ... Orthogonal lattice image data decoder 12 ... Square Lattice resampling calculation unit 13 ... square lattice image processing unit 40 ... quantization table

Claims (12)

An orthorhombic grid resampling means for resampling an imaging signal to grid points of an orthorhombic grid to generate an orthorhombic grid pixel having a luminance signal and a color signal;
An orthorhombic lattice image processing means that performs image processing on the orthorhombic lattice pixels to generate an orthorhombic lattice image signal;
An orthorhombic lattice image data compression unit that generates orthorhombic lattice image compressed data by applying image compression that lowers the contrast resolution of at least the high-frequency component of the luminance signal to the orthorhombic lattice image signal;
An image compression apparatus comprising:   The orthorhombic grid resampling means resamples the imaging signal to the grid points that are approximately half the number of imaging elements of the imaging unit that generates the imaging signal, and generates the orthorhombic grid pixels. The image compression apparatus according to claim 1. An orthorhombic lattice image data decoding unit that decodes the orthorhombic lattice image compressed data generated by the orthorhombic lattice image data compressing unit to generate orthorhombic lattice image data;
A square grid resampling means for resampling the rhombic grid image data to grid points of a square grid to generate square grid pixels;
A square lattice image processing means for performing image processing on the square lattice pixels to generate a square lattice image;
The image compression apparatus according to claim 1, further comprising:   The image compression apparatus according to claim 3, wherein the number of pixels of the square lattice pixel is substantially equal to the number of image pickup elements of an image pickup unit that generates the image pickup signal. The orthorhombic lattice resampling means generates a luminance signal and a color signal of the orthorhombic lattice pixel for each lattice point of the orthorhombic lattice based on the imaging signal,
The orthorhombic lattice image processing means adjusts luminance gradation of the luminance signal,
4. The square lattice image processing unit performs correction of a modulation degree characteristic with respect to a luminance frequency of the square lattice image data and / or correction of a contour characteristic of the square lattice image data. 5. The image compression apparatus according to 4. The orthorhombic lattice image data compression means includes a quantization table defining a reference value 0 of the orthorhombic lattice image compressed data,
The square lattice image processing means returns noise information of the square lattice image to the quantization table,
6. The image compression apparatus according to claim 5, wherein the orthorhombic lattice image data compression unit changes the definition of the reference value 0 of the quantization table based on the noise information.   The orthorhombic lattice image data compressing means is configured to form the orthorhombic lattice image signal as H pixels on a horizontal auxiliary line (where H is a natural number) × V pixels on a vertical auxiliary line (where V is a natural number) 2. 7. The rhomboid image compression data is generated independently in each unit block image after image decomposition into unit block images each consisting of × H × V pixels. The image compression apparatus as described in any one of Claims.   The orthorhombic lattice image data compression means generates the orthorhombic lattice image compressed data using a high-speed arithmetic circuit constituted by addition / subtraction and bit shift after integer conversion or fixed point conversion of the orthorhombic lattice image signal. The image compression apparatus according to claim 1, wherein the image compression apparatus is an image compression apparatus. The image compression apparatus according to any one of claims 1 to 8,
Imaging means for photoelectrically converting an optical image for each imaging device to generate the imaging signal;
An orthorhombic lattice image recording means for recording the rhombic lattice image compressed data;
An imaging recording apparatus comprising: The imaging means includes a green filter, a red filter, and a blue filter configured by a Bayer array composed of a square lattice.
The image recording apparatus according to claim 9, wherein the position of the image sensor on which the green filter is disposed is a lattice point of the rhombic lattice.   The orthorhombic lattice resampling means resamples the imaging signal from the imaging device in which the red filter and the blue filter are arranged to lattice points of the orthorhombic lattice, and for each lattice point of the orthorhombic lattice. The image recording apparatus according to claim 10, wherein a luminance signal and a color signal of the rhombic lattice pixel are generated. Re-sampling the imaging signal to the grid points of the orthorhombic grid to generate an orthorhombic grid pixel having a luminance signal and a color signal;
Applying image processing to the orthorhombic lattice pixels to generate an orthorhombic lattice image signal;
Generating an orthorhombic lattice image compression data by applying image compression that lowers the contrast resolution of at least the high-frequency component of the luminance signal to the orthorhombic lattice image signal;
An image compression program for causing a computer to execute.

Images