JP2006295521A - Moving image compression apparatus and moving image pickup apparatus - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a moving image compression apparatus capable of suitably compressing a moving image in real time in accordance with a scene, and a moving image pickup apparatus. <P>SOLUTION: A frame judgment part 109 outputs a moving image signal of a second frame rate which consists of frames picked up (extracted) at a fixed interval out of frames of an inputted moving image signal of a first frame rate. Further the frame judgment part 109 outputs a moving image signal consisting of frames (operation frames) other than the frames extracted at the fixed interval to a difference adding part 117. The moving image signal indicates a difference between an individual block of the operation frame and a block allowed to correspond by a motion vector. The difference adding part 117 adds the absolute value of the difference. A quantization table selection part 119 selects a quantization table on the basis of the added result (total value) of the difference adding part 117. The selected quantization table is used for the quantization of the moving image signal of the second frame rate. <P>COPYRIGHT: (C)2007,JPO&INPIT

Description

  The present invention relates to a moving image compression apparatus that compresses a moving image signal. The present invention also relates to a moving image imaging apparatus that includes a main moving image compression apparatus and an imaging element and performs imaging and compression of moving images.

With regard to video compression methods, various encoding methods have been planned and standardized for the following purposes, and are used in broadcasting and communication applications.
MPEG-1: Encodes video with quality exceeding the allowable limit at around 1.5Mb / s.
MPEG-2: Encodes high-quality video at around 16 Mb / s.
MPEG-4: Realizes higher efficiency, higher data operability, content and usage environment than MPEG-2.

In each compression method, in a normal case, once the compression rate is set, encoding is performed at the same compression rate even in various scenes. As a result, a redundant image may be generated, or an image deteriorated due to excessive compression may occur. Patent Document 1 describes a method of changing a compression encoding method so that distortion is not easily generated by a relatively simple calculation. In this method, an image difference from the immediately preceding frame is obtained, and the coding method is switched according to the amount of difference, large, medium, or small.
JP 2004-266720 A

  In video compression, as described above, once the compression rate is set, encoding is performed at the same compression rate in any scene, so redundant images are generated or images deteriorated due to excessive compression. There was a problem that may occur. When the compression rate is fixed, the image quality is improved if the compression rate is reduced. However, if the compression rate is reduced more than necessary, the amount of data increases and the cost of the entire system increases. The method described in Patent Document 1 is effective in improving the compression rate by simple processing, but has a problem that the compression rate cannot be determined in detail according to the scene.

  In video compression, a comparison operation between image blocks is performed to obtain a motion vector that associates an image block obtained by dividing a compression target frame with an image block obtained by dividing a reference frame based on motion in the image. The amount of calculation becomes enormous. Therefore, since high-speed arithmetic processing is required, there is a problem that the realization cost is high. The amount of calculation of the comparison operation is enormous. It is necessary to divide image data constituting one frame into a plurality of blocks, and to calculate a motion vector between the frames by comparing with each reference block. This is because the comparison operation between the blocks is performed while shifting the dots within a certain fixed range.

  The present invention has been made in view of the above-described problems, and a first object thereof is to provide a moving image compression apparatus and a moving image imaging apparatus capable of performing appropriate compression in real time according to a scene. And A second object of the present invention is to provide a moving image compression apparatus and a moving image imaging apparatus that can reduce the amount of calculation for obtaining a motion vector and perform compression at low cost.

  The present invention has been made to solve the above-described problems. The invention according to claim 1 is directed to a moving image signal input means for inputting a moving image signal having a first frame rate, and the moving image signal input. Means for extracting a frame of the moving image signal of the first frame rate inputted from the means at a constant interval and generating a moving image signal of the second frame rate; and a moving image of the first frame rate Using the signal, a selection means for selecting a quantization table used for compression of the moving image signal of the second frame rate, and the second frame rate using the quantization table selected by the selection means And a compression means for compressing the moving image signal.

  According to a second aspect of the present invention, in the moving picture compression apparatus according to the first aspect, the signal generating means extracts frames extracted at regular intervals from among the frames of the moving picture signal of the first frame rate. A moving image signal having a second frame rate is output to the compression unit, and frames other than the frames extracted at the predetermined interval are output to the selection unit.

  According to a third aspect of the present invention, in the moving picture compression apparatus according to the first aspect, a plurality of pieces of image data constituting a frame of the moving picture signal of the first frame rate input from the moving picture signal input means are provided. First block dividing means for dividing the image data into a plurality of image blocks, second block dividing means for dividing the image data constituting the reference frame into a plurality of image blocks, the image block of the compression target frame, and the reference frame The image processing apparatus further includes motion vector calculation means for calculating a motion vector for associating the image block with the motion in the image.

  According to a fourth aspect of the present invention, in the moving image compression apparatus according to the third aspect, the selection unit adds a difference between the image blocks associated by the motion vector within one frame, and the difference A difference adding means for calculating the sum of the two, and a quantization table used for compression of the compression target frame based on the sum of the differences calculated by the difference adding means from a previously recorded quantization table And a quantization table selecting means.

  According to a fifth aspect of the present invention, an image sensor that outputs an image signal in parallel from each of a plurality of divided areas that constitute an imaging region and a plurality of the image signals output from the image sensor are combined, Signal synthesizing means for outputting as a moving image signal of the frame rate, moving image signal input means for inputting the moving image signal of the first frame rate, and the first frame inputted from the moving image signal input means The second frame is extracted using signal generation means for extracting a frame of a moving image signal of a rate at a constant interval and generating a moving image signal of a second frame rate, and the moving image signal of the first frame rate. A second frame rate moving unit using a selecting unit that selects a quantization table used for compressing the moving image signal of the rate, and the quantization table selected by the selecting unit; A moving picture imaging apparatus and a compressing means for compressing the image signal.

  According to the present invention, by using a frame having a first frame rate that is higher than the frame rate of the video output and selecting a quantization table for a frame having the second frame rate that is the video output, The effect that the appropriate compression according to it can be performed in real time is acquired. According to the present invention, the calculation range in matching between the image block of the compression target frame and the image block of the reference frame is narrowed by obtaining a motion vector necessary for compression using the frame of the first frame rate. Therefore, the calculation amount for obtaining the motion vector can be greatly reduced, and compression can be performed at low cost.

  The best mode for carrying out the present invention will be described below with reference to the drawings. FIG. 1 is a block diagram showing a configuration of a moving image compression apparatus according to the first embodiment of the present invention. First, the function of each component in the figure will be described. In the moving image compression apparatus 1, the A / D conversion unit 101 (moving image signal input means) samples and quantizes a video signal (moving image signal) input at a predetermined frame rate (first frame rate). , Convert to digital data. The video memory 102 stores image data for one frame output from the A / D conversion unit 101.

  The block cutout unit 103 (first block dividing unit) converts the entire image in the frame constituting the image data output from the video memory 102 into a block (image block) of N × M pixels (N and M are natural numbers). Cut out (divide). The block extraction unit 104 sequentially extracts the blocks extracted by the block extraction unit 103 and the block extraction unit 105 (second block dividing unit) for each block. The block cutout unit 105 cuts out (divides) the entire image in the reference frame stored in the video memory 106 into N × M pixel blocks (image blocks). The video memory 106 stores image data of reference frames.

  The motion compensation prediction unit 107 (motion vector calculation means) calculates a motion vector for each block using the compression target frame and the reference frame. When calculating a motion vector, the motion compensation prediction unit 107 performs block matching within a limited range around each block after division, using an image frame divided into N × M pixel blocks, Find the motion vector. FIG. 2 shows this state. A state of block division of the image frame is as shown in FIG. For example, the motion compensation predicting unit 107 performs block matching within a range shifted by a fixed amount of dots (for example, + -15 dots) around the same coordinate position of the block a in the reference frame corresponding to the coordinate position of the block A in FIG. (See FIG. 3). The position is shifted one dot at a time, and the motion vector is obtained from the position where the absolute value difference between the pixel values of the blocks is minimized. Correspondence for each block is performed by this motion vector.

  The calculation unit 108 is a processing unit that calculates an image difference. The calculation unit 108 obtains a difference between each block divided by the block cutout unit 103 and a block associated with the motion vector obtained by the motion compensation prediction unit 107. If it demonstrates using FIG. 2, the arrow 200 has shown the motion vector calculated | required by the motion compensation prediction part 107. FIG. This portion is assumed to be an area where the difference between pixel values (the difference between the pixel values of the block A and the block b to which movement is associated by a motion vector) is minimized, and is associated by movement or the like. The calculation unit 108 obtains a difference between individual pixel values between the areas obtained in this way (block A and block b associated with movement by a motion vector), and a moving image of a difference image composed of the difference between pixel values. The image signal is output to the frame determination unit 109.

  The frame determination unit 109 (signal generation means) determines frame processing. The frame of the first frame rate input to the moving image compression apparatus 1 includes a frame used for the purpose of calculating a compression rate (quantization table) and a motion vector (hereinafter referred to as an operation frame), and includes a second frame. The rate frame is a frame that is output as a video frame. The frame determination unit 109 determines whether the processing target frame is a video frame or a calculation frame based on the sequential number of the frame. The frame determination unit 109 determines a frame that is picked up (extracted) at regular intervals based on the sequential number as a second frame rate frame (video frame) among the frames of the input video signal having the first frame rate. The moving image signal of the frame is output to the DCT unit 110. Further, the frame determination unit 109 outputs a moving image signal of a frame other than the frames extracted at regular intervals to the difference addition unit 117.

  Here, a method of processing a video frame and a calculation frame when the first frame rate is a high frame rate will be described. FIG. 4 shows a first frame rate frame (FIGS. 4B and 4C) and a second frame rate video frame (FIG. 4A) generated from the first frame rate frame. Yes. FIG. 4A shows that the video frame is 60 frames / second as an example. FIG. 4B shows a state in which frames are generated at a high frame rate of 120 frames / second. FIG. 4C shows a frame at a high frame rate equal to or higher than the frame rate in FIG. It shows how it is being generated.

  If the video output rate is 60 frames / second as shown in FIG. 4A, the frames (1), (3), and (5) in FIG. 4B are used as video frames. Frames (2) and (4) are used to calculate motion vectors and to obtain the compression ratio (quantization table) of frames (3) and (5). More specifically, the frame (2) is used to obtain a motion vector with the frame (1) as a reference frame. When a frame having a high frame rate is used, the frame interval is shortened, so that the moving distance when there is a moving object is reduced. Therefore, since it is only necessary to perform block matching only in a limited range in the surroundings, the time required for calculating a motion vector is reduced. For example, when obtaining a motion vector between frame (1) and frame (3), between frame (1) and frame (2), and between frame (2) and frame (3), respectively. When the movement distance is reduced by obtaining the motion vector, the latter requires less calculation time.

  The frame (2) is also used to calculate the absolute value difference between the blocks associated by the motion vector and to obtain a quantization table used for compression of the frame (3). This quantization table is useful for compressing the frame (3) at an appropriate compression rate according to the scene. Similarly, the frame (4) is used to calculate a motion vector and to obtain a quantization table used for compression of the frame (5).

  FIG. 4C shows the case where the first frame rate is a higher frame rate. Frames (1), (4), and (7) are used as video frames, and frames (2), (3), (5), and (6) are used as operation frames. The frame (2) or the frame (3) can be used to obtain the compression ratio (quantization table) of the frame (4). However, if the processing time is in time, the frame (3) is better than the frame (2). It is desirable to use This is because the amount of change between frames (1) and (4) is closer to using frames (1) and (3) than using frames (1) and (2). When the frame (3) is input, the frame (1) is stored in the video memory 106, and the frame (1) is used for setting the quantization table together with the frame (3). Both frames (5) and (6) can be used to determine the compression rate (quantization table) of frame (7), but using frame (6) and frame (4) in the same manner as above. A quantization table is set.

  Returning to FIG. 1, the DCT unit 110 and the quantization unit 111 (compression unit) are processing units that process video frames. The DCT unit 110 converts the signal of the difference image indicating the difference value for each block calculated by the calculation unit 108 into a signal in the spatial frequency domain (DCT conversion). This processing is the same as general MPEG processing, and Chapter 5 (MPEG Video Coding) of the literature ("Comprehensive Multimedia Selection, MPEG Video Information Media Society", published by Ohm Publishing, April 20, 1996). )It is described in.

  The quantization unit 111 quantizes the DCT coefficient using the quantization table 120 selected by the quantization table selection unit 119, and performs image compression. The quantization table 120 is obtained by the processing using the above-described calculation frame, and is set so that the quantization step is finely cut, the compression rate is small, and the image quality is improved in a scene with motion. Conversely, in a scene where there is no motion, the value of the quantization step is large, and the compression rate is set high within a range that does not affect the image quality.

  In addition, the quantization unit 111 generates a copy of the output. One of the copy output and the copy is output as video output encoded data via the variable length encoding unit 115 and the buffer 116. The other is stored in the video memory 106 via the inverse quantization unit 112, the inverse DCT unit 113, and the calculation unit 114, and is used as a reference frame for obtaining a difference between frames at the time of compression.

  The inverse quantization unit 112 performs inverse quantization (expansion) on the frame quantized by the quantization unit 111. The inverse DCT unit 113 decodes the frequency data. The calculation unit 114 is a processing unit that performs a process opposite to the process in which the calculation unit 108 obtains a difference between frames, and the difference data output from the inverse DCT unit 113 and the data divided by the block cutout unit 105. Are added to restore the original frame data. The frame data restored by the calculation unit 114 is stored in the video memory 106. In this way, the video frame is returned to the original image data, and is used as a reference frame when obtaining a motion vector between frames.

  The variable length encoding unit 115 is a processing unit that outputs a video frame as video data, and compresses (encodes) the video frame. The variable length coding unit 115 performs variable length coding by a method of assigning a short code to data having a higher appearance frequency. The buffer 116 is storage means for buffering output encoded data. The video data stored in the buffer 116 is sequentially extracted as encoded data and output.

  The difference addition unit 117 (difference addition unit), the addition value storage unit 118, the quantization table selection unit 119 (quantization table selection unit), and the quantization table 120 include a quantization table using a moving image signal of an operation frame. Related to selection and setting. The difference addition unit 117 receives the moving image signal of the calculation frame distributed by the frame determination unit 109 for each block. This moving image signal indicates a difference between each block of the calculation frame and a block associated with the motion vector. The difference adding unit 117 adds the absolute values of the differences. The result of adding the difference is stored in the added value storage unit 118 every time the difference of each block is added. A total value in the entire frame, which is obtained by adding the differences of all the blocks constituting one frame, is used to determine the magnitude of motion between scenes. The difference addition unit 117 outputs the total value to the quantization table selection unit 119.

  The quantization table selection unit 119 selects a quantization table based on the addition result (total value) by the difference addition unit 117. The total value output from the difference adding unit 117 indicates a change in the moving image signal based on the movement of the subject. When the total value is large (for example, when the total value is greater than or equal to a predetermined value), the motion in the image is large, and the quantization table selection unit 119 selects a quantization table with a small quantization step. Conversely, when the total value is small (for example, when the total value is less than the predetermined value), the motion in the image is small, and the quantization table selection unit 119 selects a quantization table with a large quantization step. To do. Thus, an appropriate quantization table corresponding to the scene is selected. The quantization table selection unit 119 sets the selected quantization table as the quantization table 120 in the quantization unit 111 and the inverse quantization unit 112.

  Next, the operation of the moving image compression apparatus 1 according to the present embodiment will be described using the flowcharts shown in FIGS. The video signal input to the moving image compression apparatus 1 is converted into digital data by the A / D converter 101 and stored in the video memory 102 (step S501). The block cutout unit 103 divides the image data output from the video memory 102 into N × M pixel blocks, and the block cutout unit 105 converts the reference frame image data output from the video memory 106 into N × M pixel blocks. Divide into blocks (step S502). The block extraction unit 104 sequentially extracts the blocks divided by the block cutout unit 103, and extracts a necessary block from the reference frame blocks divided by the block cutout unit 105 (step S503).

  Thereafter, when block extraction is completed (YES in step S504), processing in step S511 is performed. This means that the processing of the last block of one frame has been completed. On the other hand, if the block extraction has not been completed (NO in step S504), the process of step S505 is performed. Which process is to be performed is determined by, for example, a control unit (not shown), and the processing unit corresponding to each process is controlled.

  In step S505, the motion compensation prediction unit 107 obtains a motion vector using the extracted block and the block of the reference frame. As shown in FIG. 2, the motion compensation prediction unit 107 stores each block as a reference frame in the video memory 106 while shifting the pixel position within an area including a certain range centering on the current corresponding position. Block matching with existing image data. The motion compensation prediction unit 107 sequentially shifts the pixel position, and obtains a motion vector by regarding the position where the absolute value difference between the two pixel values for which block matching is performed as a minimum, as the area where the block has moved.

  Subsequently, the calculation unit 108 obtains a difference value between the blocks associated with the motion vector (step S506). The frame determination unit 109 determines whether or not the processing target frame is a video frame based on the frame counter value, and performs frame distribution (step S507). When it is determined that the processing target frame is a video frame, the frame determination unit 109 outputs a moving image signal indicating the difference value obtained by the calculation unit 108 to the DCT unit 110. The DCT unit 110 performs DCT on the difference value and converts it into DCT coefficients in the spatial frequency domain (step S508).

  The quantization unit 111 quantizes the DCT coefficient using the quantization step specified in the quantization table 120 (step S509). If the quantization step size is reduced, the image quality is improved, but the amount of data increases, and if the quantization step size is increased, the quantization step has the property that the image quality deteriorates and the data amount decreases. Yes. After step S509, the process returns to step S503, and the next block is processed.

  On the other hand, if it is determined in step S507 that the processing target frame is not a video frame, processing of a calculation frame captured at a high frame rate (step S510) is performed. In step S <b> 510, the frame determination unit 109 outputs a moving image signal indicating the difference value obtained by the calculation unit 108 to the difference addition unit 117. The difference addition unit 117 converts the difference value into an absolute value and adds it to the addition value up to the previous block. This added value is stored in the added value storage unit 118. Subsequent processing returns to the processing of step S503.

  Note that when there are a plurality of computation frames between adjacent video frames, the processing of step S510 is performed in each computation frame, but the following may be performed. For example, after the processing shifts to the processing of the next computation frame, the addition value of the previous computation frame stored in the addition value storage unit 118 is erased and then the addition value of the next computation frame is written. do it. As a result, the quantization table of frame (4) in FIG. 4C is selected based on the sum of the difference values of frame (3), not frame (2). Alternatively, based on the sequential number of the frame, the difference addition process for the calculation frames other than the calculation frame having the preset sequential number may be skipped.

  The process of step S511 is a process when the process of the last block of one frame is completed. Thereafter, when the processing target frame is a video frame (YES in step S511), the process of step S512 is performed. If the processing target frame is a calculation frame (NO in step S511), the subsequent processing is processing in step S514.

  In step S512, the quantization unit 111 creates a copy of the frame, outputs one of the original frame and its copy to the inverse quantization unit 112, and outputs the other to the variable length coding unit 115. Thereafter, two processes for processing a video frame are performed in parallel (step S513). These processes are shown in FIGS. 6 (a) and 6 (b). In FIG. 6A, the variable length encoding unit 115 encodes quantized data by assigning shorter data to data having a higher appearance frequency (step S601). The variable length encoding unit 115 stores the encoded data in the buffer 116 (step S602). This data is output as encoded data.

  In FIG. 6B, since it is necessary to use a video frame later as a motion compensation reference frame, a process of decoding the video frame into an original image frame and storing it in the video memory 106 is performed. The inverse quantization unit 112 performs inverse quantization of the video frame using the quantization table 120, and outputs the video frame to the inverse DCT unit 113 (step S603). The inverse DCT unit 113 performs inverse DCT transform on the inversely quantized video frame, and outputs the video frame to the arithmetic unit 114 (step S604). The calculation unit 114 adds information other than the difference of one frame output from the block cutout unit 105 to the video frame output from the inverse DCT unit 113 and decodes it to the original image frame (step S605). Then, it is stored in the video memory 106 as a reference frame (step S606).

  Returning to FIG. 5, the processing after step S514 is processing of the calculation frame. The difference addition unit 117 outputs a total value obtained by adding the absolute values of the differences between the blocks for one frame to the quantization table selection unit 119. The quantization table selection unit 119 selects the quantization table 120 based on the total value. If the total value is large, it is assumed that the movement is intense, so the quantization table selection unit 119 selects a quantization table with a small quantization step. If the total value is small, it is assumed that there is not much movement, and the quantization table selection unit 119 selects a quantization table with a large quantization step (step S514).

  The quantization table selection unit 119 sets the selected quantization table as the quantization table 120 in the quantization unit 111 and the inverse quantization unit 112 (step S515). The quantization table 120 is used for quantization and inverse quantization of the next video frame. A plurality of quantization tables are prepared, and a quantization table suitable for the quantization step is selected from them according to the total size of the absolute values of the differences.

  As described above, according to the present embodiment, the quantization used for compression of the frame at the second frame rate that is the video output using the frame at the first frame rate that is higher than the frame rate of the video output. By selecting the table, appropriate compression according to the scene can be performed in real time. In addition, when the motion between frames becomes large, the quantization step is set small, and as a result, the compression rate is reduced. Therefore, it is possible to prevent image deterioration due to excessive compression. In particular, it is possible to greatly reduce image degradation due to an inappropriate compression rate even in a scene where a leaf or a flower that is in full bloom is swaying in the wind. In addition, when the motion between frames is small, the compression rate is set high, but it is possible to avoid generating redundant images and increasing the amount of data more than necessary.

  In this embodiment, the time interval between frames of the first frame rate is set to be shorter than that of the second frame rate. By performing processing for obtaining a motion vector necessary for compression using a frame at the first frame rate, the calculation range in matching between the image block of the compression target frame and the image block of the reference frame can be narrowed. Therefore, the amount of calculation for obtaining the motion vector can be greatly reduced, and high-definition images can be compressed at low cost.

  Next, a second embodiment of the present invention will be described. FIG. 7 is a block diagram illustrating a configuration of the moving image capturing apparatus according to the present embodiment. The moving image imaging apparatus according to the present embodiment includes the moving image compression apparatus 1 according to the first embodiment, and realizes generation of a high frame rate moving image signal by reading pixel information in parallel from the image sensor. . Hereinafter, functions of configurations other than the moving image compression apparatus 1 already described will be described. The imaging lens 30 condenses the emitted light emitted from the subject. The solid-state imaging device 31 (imaging device) receives the emitted light collected by the imaging lens 30, converts it into an electrical signal, and outputs a plurality of image signals. The imaging signal processing unit 32 (signal synthesis unit) synthesizes and adjusts a plurality of image signals output from the solid-state imaging device 31, and adjusts the video signal as a first frame rate video output signal (video signal). Output to 1.

  FIG. 8 shows the internal configuration of the solid-state imaging device 31. In the solid-state imaging device 31, the imaging region is divided into a plurality of areas (divided areas), and a charge signal is extracted from each divided area and output in parallel as a plurality of output signals. The imaging area is divided into four imaging areas I, II, III, and IV. The charges in the imaging region I are read out by the vertical transfer CCD 315 and the horizontal transfer CCD 311 and output as an output a. The charges in the imaging region II are read out by the vertical transfer CCD 316 and the horizontal transfer CCD 312 and output as an output b. The charges in the imaging region III are read out by the vertical transfer CCD 317 and the horizontal transfer CCD 313 and output as an output c. The charges in the imaging region IV are read out by the vertical transfer CCD 318 and the horizontal transfer CCD 314 and output as an output d. Thus, a high frame rate is realized by dividing the imaging region and reading out pixel signals in parallel.

  Here, outputs a to d are obtained, but signals are read from all the pixel regions by all (four) signal outputs. The imaging signal processing unit 32 buffers outputs a to d in units of one frame, synthesizes them, and outputs them as video output signals. This video output signal is input to the A / D converter 101 as a video input. The processes after the A / D converter 101 are the same as those in the first embodiment. In the present embodiment, the image sensor provided with the CCD sensor has been described. However, an image sensor provided with a CMOS sensor may be used.

  As described above, according to the present embodiment, it is possible to acquire a frame at a high frame rate by reading out a pixel signal using an image pickup device that performs parallel reading, and it is easy to realize the moving image pickup apparatus of the present invention. .

  As described above, the embodiments of the present invention have been described in detail with reference to the drawings. However, the specific configuration is not limited to these embodiments, and includes design changes and the like without departing from the gist of the present invention. .

It is a block diagram which shows the structure of the moving image compression apparatus by the 1st Embodiment of this invention. FIG. 5 is a reference diagram for explaining a motion vector calculation method according to the first embodiment of the present invention. FIG. 5 is a reference diagram for explaining a motion vector calculation method according to the first embodiment of the present invention. It is a reference diagram for demonstrating the video frame and calculation frame in the 1st Embodiment of this invention. It is a flowchart which shows operation | movement of the moving image compression apparatus by the 1st Embodiment of this invention. It is a flowchart which shows operation | movement of the moving image compression apparatus by the 1st Embodiment of this invention. It is a block diagram which shows the structure of the moving image compression apparatus by the 2nd Embodiment of this invention. It is a schematic block diagram which shows the structure of the solid-state imaging device by the 2nd Embodiment of this invention.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 ... Moving image compression apparatus, 30 ... Imaging lens, 31 ... Solid-state imaging device, 32 ... Imaging signal processing part, 101 ... A / D conversion part, 102, 106 ... Video Memory, 103, 105 ... Block cutout unit, 104 ... Block extraction unit, 107 ... Motion compensation prediction unit, 108,114 ... Calculation unit, 109 ... Frame determination unit, 110 ... DCT unit, 111 ... quantization unit, 112 ... inverse quantization unit, 113 ... inverse DCT unit, 115 ... variable length coding unit, 116 ... buffer, 117 ... difference addition 118, addition value storage unit, 119 ... quantization table selection unit, 120 ... quantization table, 311, 312, 313, 314 ... horizontal transfer CCD, 315, 316, 317, 318 ... Vertical transfer CCD

Claims (5)

Moving image signal input means for inputting a moving image signal of a first frame rate;
Signal generating means for extracting a frame of the moving image signal of the first frame rate inputted from the moving image signal input means at a constant interval, and generating a moving image signal of the second frame rate;
Selecting means for selecting a quantization table to be used for compression of the moving image signal of the second frame rate using the moving image signal of the first frame rate;
Compression means for compressing the moving image signal of the second frame rate using the quantization table selected by the selection means;
A moving image compression apparatus.   The signal generating means outputs frames extracted at regular intervals out of the frames of the moving image signal at the first frame rate to the compression means as moving image signals at the second frame rate, and the regular intervals The moving image compression apparatus according to claim 1, wherein a frame other than the frame extracted in step (a) is output to the selection unit. First block dividing means for dividing image data constituting a frame of the moving image signal of the first frame rate input from the moving image signal input means into a plurality of image blocks;
A second block dividing means for dividing the image data constituting the reference frame into a plurality of image blocks;
Motion vector computing means for computing a motion vector for associating the image block of the compression target frame with the image block of the reference frame based on the motion in the image;
The moving image compression apparatus according to claim 1, further comprising: The selection means includes
A difference adding means for adding a difference between the image blocks associated by the motion vector within one frame, and calculating a sum of the differences;
A quantization table selection unit that selects a quantization table to be used for compression of a frame to be compressed, based on the sum of the differences calculated by the difference addition unit, from pre-recorded quantization tables;
The moving picture compression apparatus according to claim 3, further comprising: An image sensor that outputs image signals in parallel from each of a plurality of divided areas constituting the imaging region;
A signal combining unit that combines the plurality of image signals output from the image sensor and outputs the image signals as a moving image signal having a first frame rate;
Moving image signal input means for inputting a moving image signal of the first frame rate;
Signal generating means for extracting a frame of the moving image signal of the first frame rate inputted from the moving image signal input means at a constant interval, and generating a moving image signal of the second frame rate;
Selecting means for selecting a quantization table to be used for compression of the moving image signal of the second frame rate using the moving image signal of the first frame rate;
Compression means for compressing the moving image signal of the second frame rate using the quantization table selected by the selection means;
A moving image pickup apparatus.

Images