JP4628374B2 - Encoding device and program - Google Patents

Description

  The present invention relates to an encoding device and a program.

  As an encoding technique for compressing the information amount of moving image data, an MPEG-2 system (for example, standardized by ISO (International Organization for Standardization) / IEC (International Electrotechnical Commission) JTC1 / SC29 / WG11 (MPEG)) (for example, Non-Patent Document 1) and H.264 standardized by MPEG and ITU-T (International Telecommunication Union Telecommunication Standardization Sector). An H.264 / MPEG-4 AVC (Advanced Video Coding) system (see, for example, Non-Patent Document 2) is known.

  In any method, the moving image data is composed of a series of still images (hereinafter referred to as frames), and the above-described compression is realized by encoding each frame.

  In this encoding, each frame is divided into a plurality of blocks of a predetermined size, and each process of predictive encoding, orthogonal transform, and quantization is performed for each block. In addition, there are two types of encoding, intra encoding and inter encoding. In inter encoding, motion prediction processing is performed in addition to the above processes. Hereinafter, a frame that is a target of intra coding is referred to as an intra frame, and a frame that is a target of inter coding is referred to as an inter frame.

  When a certain block is encoded, the encoding apparatus refers to another part of the moving image data (hereinafter referred to as a prediction basic part) based on a prediction image generation method selected from several prediction image generation methods. If the target block is a block included in an intra frame, at least a part of other parts in the same frame is a prediction basic part. A prediction image is generated using at least a part of other portions in other frames as a prediction basic portion. Further, difference data indicating a difference between the prediction image and an image related to the block is generated. Obtain (predictive coding process). Then, each process of orthogonal transform and quantization is performed on the acquired difference data. Several combinations of the predicted image generation method and the selection method of the prediction basic part are determined in advance, and each is called a prediction mode.

  Here, the prediction mode concerning intra coding will be described in detail. The encoding device sequentially encodes each block constituting the frame from the upper left of the frame in the right direction, and when reaching the right end, it sequentially encodes down one step and toward the right. In intra coding, when performing predictive coding of a certain block, the coding apparatus acquires a prediction base portion from images included in other blocks that have already been coded in the same frame.

  In a more specific example, the block is usually composed of an area of 16 × 16 pixels or 4 × 4 pixels. When one block is composed of an area of 16 × 16 pixels, the upper side of the block is The encoded block adjacent to the left side is a selection target of the prediction basic part. There are four types of prediction modes, called horizontal prediction, vertical prediction, DC (direct current) prediction, and plane prediction, depending on the combination of the combination of the predicted image generation method and the selection of the prediction basic part. After temporarily generating difference data in each prediction mode, the difference data having the highest encoding efficiency is used as the output of the prediction encoding process (see, for example, FIG. 3 of Patent Document 1). ). On the other hand, when one block is composed of a 4 × 4 pixel area, in addition to the encoded blocks adjacent to the upper and left sides of the block, the encoded block diagonally above and to the right of the block is also predicted. It becomes the selection target of the basic part. The prediction mode includes horizontal direction prediction, vertical direction prediction, DC (direct current) prediction, diagonal lower left direction prediction, diagonal lower right direction prediction, vertical right direction, depending on the combination of the selection of the prediction basic part and the predicted image generation method. There are nine methods called direction prediction, horizontal downward prediction, vertical left prediction, and horizontal upward prediction. The encoding device temporarily generates difference data in each prediction mode, The one having the highest conversion efficiency is used as the output of the predictive encoding process (see, for example, FIG. 5 of Patent Document 1).

  Note that the differential data (prediction mode) determination method with the best coding efficiency includes a rate distortion optimization mode determination method using rate distortion (RD: Rate Distortion) and a mode determination method that does not use rate distortion. Are known.

  For a block of 16 × 16 pixels, a mode decision method that does not use rate distortion is used. That is, among the four types of prediction modes, the prediction mode in which the error (D: Distortion) between the predicted image and the original image (the image related to the encoding target block) is the smallest is the prediction mode with the highest encoding efficiency. It is determined.

  On the other hand, for the 4 × 4 pixel block, any of the above mode determination methods is used. When the rate distortion optimization mode determination method is used, an error (D: Distortion) between a predicted image and an original image among nine types of prediction modes and an amount of information generated when the prediction mode is selected (R: Rate) ) And D + λR, which is defined as a cost, is calculated, and the prediction mode with the smallest value is determined to be the prediction mode with the best coding efficiency. Here, λ is called a Lagrangian coefficient and is defined by a function of a quantization parameter Qp related to the quantization process. When using a mode determination method that does not use rate distortion, an amount calculated from the error between the predicted image and the original image and an amount determined based on whether or not the prediction mode has occurred in the same frame so far As a cost, it is determined that a prediction mode with a small value is the prediction mode with the best coding efficiency.

  Note that the encoding device uses a prediction mode for a 16 × 16 pixel block or encodes a prediction mode for a 4 × 4 pixel block by dividing the block into 4 × 4 pixels when encoding a 16 × 16 pixel block. Whether or not to use is also determined. In this determination, either the rate distortion optimization mode determination method or the mode determination method that does not use rate distortion is used.

  In the above, the prediction mode concerning intra coding was demonstrated.

  By the way, in the predictive encoding process, even if it is differential data for the same encoding target block, the value differs depending on which prediction mode is used. Then, the difference between the values of the difference data may appear as a large difference in the output image after decoding. This phenomenon is called flicker.

  The flicker will be described in detail. As described above, the difference data obtained as a result of the predictive encoding process is orthogonally transformed and further quantized and output, but when two difference data having different values are orthogonally transformed, Also in the transform coefficient data obtained as a result of the orthogonal transform, a difference reflecting the difference value of the difference data appears. This difference is amplified in some cases by quantization processing. Once the difference is amplified, it does not return to the original value even after decoding, so a large difference occurs in the output image after decoding.

  The problem of flicker is particularly with respect to moving image data or a portion thereof that hardly changes with time. For blocks that change greatly in time, this is not a problem because the viewer of the reproduced moving image data cannot feel flicker, but if there is almost no change in time, the viewer is sensitive to flicker. This is a big problem as image quality degradation.

  Flicker is particularly likely to occur between an intra frame and the immediately preceding frame. That is, since a block related to an intra frame is encoded using information of another part in the same intra frame, an image related to the encoding target block is almost the same as an image related to a block at the same position in a past frame. Even if it does not change, the prediction mode will be different if there are significant changes in other parts. When the prediction mode is different, a difference occurs in the difference data as described above. As a result, a large difference may occur in the output image, but since there is almost no temporal change in the portion related to the encoding target block, the observer of the reproduced moving image data expects no change. Nevertheless, a large difference has occurred in the output image, and as a result, the observer perceives the occurrence of flicker.

  Intraframes must be inserted periodically (usually at intervals of about once every 0.5 to 2 seconds) to achieve random access such as fast forward and rewind. For this reason, flicker tends to occur regularly, and it is desired to suppress the occurrence of flicker.

  The technique disclosed in Patent Document 2 is one of the techniques for suppressing the occurrence of flicker. When determining the prediction mode of each block of an intra frame, the encoding apparatus according to this technique performs not only the above-described encoding efficiency but also an evaluation that takes into consideration a difference from a past frame, and then performs prediction. Determine the mode. Specifically, the prediction mode is evaluated using time difference information (see paragraphs 0066 and 0067 of Patent Document 2) indicating a difference in the time direction.

  Non-Patent Documents 3 and 4 and Patent Documents 3 to 6 also describe techniques for suppressing the occurrence of flicker.

Note that Non-Patent Document 2 describes a process called weighted prediction. This processing is processing for realizing fade-in / fade-out at the time of encoding. Specifically, the luminance value at the start of fade-in (fade-out), the luminance value at the end of fade-in (fade-out), Is calculated by multiplying the original luminance value by the weight and then encoding the luminance value.
JP 2005-39321 A JP 2005-318468 A Japanese Patent Laid-Open No. 2005-217743 JP 2005-348008 A JP 2006-94081 A JP 2006-191287 A "ISO (International Organization for Standardization) / IEC (International Electrotechnical Commission) 13818-2", 1994 “ITU-T (International Telecommunication Union Telecommunication Standardization Sector) Rec. (Recommendation) H.264 | ISO (International Organization for Standardization) / IEC (International Electrotechnical Commission) 14496-10 Advanced Video Coding (Advanced Video Coding)”, 2005 Chono et al., “Method of reducing I-picture flicker by quantization control with detent”, PCSJ (Image Coding Symposium) 2005, 2005, P-4.02. Shimizu et al., “Flicker reduction method for intra-coded frames in H.264 / AVC”, PCSJ (Image Coding Symposium) 2005, 2005, P-4.03.

  However, in the technique disclosed in Patent Document 2, in order to determine the prediction mode after performing an evaluation taking into consideration not only the encoding efficiency but also the difference from the past frame, the determined prediction mode is When predictive encoding is used, there is a problem that the intra-frame encoding efficiency deteriorates compared to the case of using a prediction mode determined after performing an evaluation that does not take into account the difference from the past frame. is there.

  Accordingly, an object of the present invention is to provide an encoding device, an encoding method, and a program that can suppress the occurrence of flicker and also suppress the deterioration of the encoding efficiency of an intra frame.

  An encoding apparatus according to the present invention for solving the above-described problems is composed of n frames (n ≧ 3) constituting moving image data, and the n-th frame is a series of frames to be intra-encoded. Frame acquisition means for sequentially acquiring, at least a part of each pixel constituting each of the 2nd to (n-1) th frames, a pixel value of at least a part of each pixel constituting the first frame, and the nth Based on at least a part of pixel values of each pixel constituting the frame, pixel value determining means for determining the pixel value, and using the pixel values determined by the pixel value determining means, the second to n And encoding means for encoding each first frame.

  According to this, based on the pixel value of the pixel constituting the first frame and the pixel value of the pixel constituting the nth frame, the pixel value of the pixel constituting each of the second to (n-1) th frames is determined. Since it can be determined, for example, the interval between them can be smoothly connected, and the occurrence of flicker can be suppressed in the nth frame which is an intra frame. In addition, since the flicker generation suppressing process does not require a process for changing the prediction mode of the nth frame, it is possible to suppress the deterioration of the intra-frame coding efficiency.

  In the encoding apparatus, the pixel value determining unit may include one of the pixels constituting the first frame with respect to one of the pixels constituting the second to (n-1) th frames. The pixel value may be determined based on an interpolation line formed by interpolating one pixel value and one pixel value of each pixel constituting the nth frame.

  According to this, it becomes possible to smoothly connect between the pixel value of the pixel constituting the first frame and the pixel value of the pixel constituting the nth frame by interpolation.

  In each of the encoding devices, a dividing unit that divides the n-th frame into a plurality of blocks and a target block that is one of the blocks for each of the first to n−1th frames. A corresponding part acquiring means for acquiring a part corresponding to the at least one pixel value, and the pixel value determining means includes at least a part of pixel values of each pixel constituting the target block and the corresponding part of the first frame. Based on at least some pixel values of each pixel constituting the part acquired by the acquisition means, each part acquired by the corresponding part acquisition means among the second to (n-1) th frames is configured. It is good also as determining the pixel value of at least one part of each pixel to perform.

  According to this, the pixel for determining the pixel value can be appropriately determined based on the correspondence between the block of interest and each part.

  Further, in this encoding apparatus, the image content of the block of interest, the image content of each part acquired for at least a part of each of the first to n−1 frames by the corresponding part acquisition unit, A similarity calculation means for calculating the similarity of the image, and a determination means for determining whether or not to perform pixel value determination by the pixel value determination means for the block of interest according to the similarity calculated by the similarity calculation means And the pixel value determining means includes at least a part of each pixel constituting the target block and the first value when the pixel value is determined by the determining means to determine the pixel value for the target block. And from the second to the (n−1) th frame based on at least a part of each pixel constituting the portion acquired by the corresponding portion acquisition means Determining at least part of the pixel values of the pixels constituting the respective portions obtained by the corresponding portion acquisition unit of the frame, it is also possible.

  According to the present invention, the pixel values of the pixels constituting each of the second to (n-1) th frames are set to the pixel value of the pixels constituting the first frame and the pixel value of the pixels constituting the nth frame. Therefore, in some cases, the pixel value may be far from the original pixel value. According to the above encoding device, whether or not to determine the pixel value by the pixel value determination unit is determined based on the similarity between the target block and at least a part of each corresponding part. It is possible to prevent the pixel value determination by the pixel value determination means. As a result, it is possible to prevent the pixel values of the pixels constituting the second to (n-1) th frames from being far from the original pixel values.

  Further, in each of the encoding devices, the corresponding portion acquisition unit may include a portion at the same position as the position of the block of interest in the nth frame in each of the first to (n−1) th frames. It is good also as acquiring as a part corresponding to the said attention block.

  According to this, as a corresponding portion, a portion at the same position as the target block can be used in each frame.

  Further, in each of the encoding devices, the corresponding part acquisition unit acquires a motion vector indicating a position in each of the first to (n-1) th frames for the content of the image included in the block of interest. A vector acquisition unit may be included, and a portion corresponding to the block of interest may be acquired for each of the first to (n-1) th frames based on the motion vector.

  According to this, even when the image included in the block of interest moves over time in the frame, the corresponding part acquisition unit can acquire an appropriate corresponding part. .

  The program according to the present invention is a frame acquisition means for sequentially acquiring n (n ≧ 3) frames constituting moving image data, and at least one of the pixels constituting each of the 2nd to (n-1) th frames. Pixel values for determining pixel values based on at least some pixel values of each pixel constituting the first frame and at least some pixel values of each pixel constituting the nth frame A program for causing a computer to function as a determination unit and a prediction encoding unit that predictively encodes each of the second to (n-1) th frames using the pixel value determined by the pixel value determination unit. .

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

  FIG. 1 is a schematic block diagram showing a system configuration and functional blocks of an encoding apparatus 1 according to the present embodiment. As shown in the figure, the encoding apparatus 1 includes an intra / inter image determination unit 10, a macroblock division unit 12, a frame delay unit 14, a corresponding portion acquisition unit 16, a processing target determination unit 18, and a normal encoding processing unit 20. , An output unit 22, an intra image frame memory unit 24, an inter image frame memory unit 26, an inter encoding target image setting unit 28, and an inter image encoding unit 30. The corresponding portion acquisition unit 16 includes an intra macroblock similarity comparison unit 160 and an inter macroblock similarity comparison unit 161 therein, and the inter coding target image setting unit 28 includes a pixel value therein. An acquisition unit 280 and a pixel value determination unit 281 are included.

  The intra / inter image determination unit 10 sequentially acquires a series of frames (input images) constituting moving image data. Then, each frame is classified into either an intra frame or an inter frame. That is, for each frame, it is determined which of the intra coding and inter coding is to be used.

  In one example, the intra / inter image determination unit 10 classifies the acquired frames into intra frames at a predetermined cycle. In this case, an intra frame insertion period is preset in the intra / inter image determination unit 10. The intra / inter image determination unit 10 determines whether each frame is a target of intra coding or a target of inter coding based on the set intra frame insertion period.

  In another example, the intra / inter image determination unit 10 detects that the image content has greatly changed (scene change or the like) between two consecutive frames (referred to as a first frame and a second frame). I do. When the change is detected, the intra / inter image determination unit 10 determines that the second frame is a target of intra encoding. The intra / inter image determination unit 10 calculates an average value within a frame of pixel values (values indicating luminance or color difference) of each pixel constituting each frame, and performs the above detection based on the average value within the frame. It is preferable to do so.

  In the following, the description will be focused on n frames (n ≧ 3) out of a series of frames acquired by the intra / inter image determination unit 10. Here, of the n frames, the first frame and the nth frame are both intra frames, and the second to (n-1) th frames are all inter frames.

  The macroblock division unit 12 divides each frame classified into either an intra frame or an inter frame by the intra / inter image determination unit 10 into a plurality of blocks. Then, the divided frames are sequentially output to the frame delay unit 14. Here, it is assumed that this block is a block (referred to as a macro block) having a size of 16 × 16 pixels. However, this is for convenience only and is not limited to this size.

  The frame delay unit 14 sequentially receives input of each frame from the macroblock division unit 12. Then, for each input frame, the intra / inter image determination unit 10 determines whether the frame is classified into an intra frame or an inter frame. When it is determined that a frame input at a certain time is classified as an intra frame, the frame delay unit 14 converts each frame (image) input until the next intra frame is input into an inter image. The data is stored in the frame memory unit 26. Further, the frame delay unit 14 sequentially outputs each frame determined to be classified as an intra frame to the corresponding part acquisition unit 16.

  Here, the frame delay unit 14 outputs the first frame and the n-th frame to the corresponding portion acquisition unit 16 and causes the inter-image frame memory unit 26 to store the second to n−1th frames.

  The corresponding part acquisition unit 16 sequentially pays attention to each of the blocks constituting the nth frame input from the frame delay unit 14. Then, for each of the first to n−1th frames, a portion corresponding to the block of interest is acquired. It is desirable that this part has the same shape as the target block.

  In a specific example, the corresponding part acquisition unit 16 corresponds to a part corresponding to a target block in each of the first to (n−1) th frames at a position that is the same as the position of the target block in the nth frame. Get as.

  FIG. 2 is a diagram illustrating a specific example of the corresponding part acquired by the corresponding part acquisition unit 16 in this example. In the figure, an example of n = 4 is shown. The frames F1, F2, F3, and F4 are the first, second, third, and fourth frames, respectively, the frames F1 and F4 are intra frames, and the frames F2 and F3 are inter frames. In addition, the target block B is shown in the frame F4.

  As shown in FIG. 2, the corresponding part acquisition unit 16 corresponds to the target block B with the parts P1a, P2a, and P3a at the same position as the target block B in the frame F4 in each of the frames F1, F2, and F3. Get as part to do.

  In another example, the corresponding part acquisition unit 16 acquires a motion vector indicating the position in each of the 1st to (n-1) th frames of the content of the image included in the block of interest. Then, based on the acquired motion vector, a portion corresponding to the block of interest is acquired for each of the first to n−1th frames.

  FIG. 3 is a diagram illustrating a specific example of the corresponding part acquired by the corresponding part acquisition unit 16 in this example. This figure shows an example of n = 4 as in FIG. The frames F1, F2, F3, and F4 are the first, second, third, and fourth frames, respectively, the frames F1 and F4 are intra frames, and the frames F2 and F3 are inter frames. In addition, the target block B is shown in the frame F4.

  As illustrated in FIG. 3, the corresponding portion acquisition unit 16 acquires a motion vector indicating the positions C1, C2, and C3 in the frames F1, F2, and F3 for the content C4 of the image included in the block of interest B. The motion vector acquired in this way represents the amount of motion from the position in each frame when the image content C4 included in the block of interest B moves within the frame in terms of time. The corresponding part acquisition unit 16 acquires the parts P1b, P2b, and P3b corresponding to the block of interest B for each of the frames F1, F2, and F3 based on the acquired motion vector.

  The intra macroblock similarity comparison unit 160 calculates the similarity between the image content of the block of interest and the image content of the part acquired for the first frame by the corresponding part acquisition unit 16. In this case, the intra macroblock similarity comparison unit 160 acquires the image content of the first frame based on the image stored for the first frame in the intra image frame memory unit 24 described later. It is desirable.

  Hereinafter, a specific example of similarity calculation will be described. In one example, the intra macroblock similarity comparison unit 160 calculates, for each pixel, the difference between the pixel value of each pixel in the target block and the pixel value of each pixel in the corresponding portion. Then, the similarity is calculated by comparing the sum with a predetermined threshold. In this case, the similarity indicates dissimilarity when the sum is equal to or greater than the threshold, and indicates similarity when the sum is less than the threshold.

  In another example, the intra macroblock similarity comparison unit 160 calculates the sum of the pixel values of each pixel in the target block and the sum of the pixel values of each pixel in the corresponding portion, and further calculates the difference. Is calculated. Then, the similarity is calculated by comparing the calculated difference with a predetermined threshold. In this case, the similarity indicates dissimilarity when the difference is greater than or equal to a threshold, and indicates similarity when the difference is less than the threshold.

  The inter macroblock similarity comparison unit 161 calculates the similarity between the image content of the block of interest and the image content of the portion acquired for each of the second to n−1 frames by the corresponding portion acquisition unit 16. Specifically, the inter-macroblock similarity comparing unit 161 calculates the similarity for each of the second to n−1th frames by the same process as the similarity calculating process by the intra-macroblock similarity comparing unit 160. calculate. In this case, the intra macroblock similarity comparison unit 160 acquires the image contents of the second to n−1th frames based on the image stored in the inter image frame memory unit 26. It is desirable to do.

  The processing target determination unit 18 is a pixel value determination unit to be described later for each block constituting the nth frame according to the similarity calculated by the intra macroblock similarity comparison unit 160 and the inter macroblock similarity comparison unit 161. It is determined whether or not the pixel value is determined according to H.281.

  In other words, the processing target determination unit 18, when there is one for which a similarity degree indicating dissimilarity is calculated in any one of the first to n−1th frames for a certain block, will be described later. It is determined that the pixel value determination unit 281 does not determine the pixel value. On the other hand, when the similarity indicating similarity is calculated for all of the first to (n−1) th frames for a certain block, the processing target determination unit 18 uses a pixel value determination unit 281 described later for the block. It is determined that the pixel value is determined.

  The normal encoding processing unit 20 performs an encoding process including the above-described prediction encoding, orthogonal transform, and quantization processes on all intra frames for each block. In addition, for the block including the portion corresponding to the block determined not to determine the pixel value by the processing target determination unit 18 in the inter frame, the prediction encoding, the orthogonal transformation, the quantization, and the motion prediction described above are performed. Encoding processing including each processing is performed. When the encoding process is completed, the normal encoding processing unit 20 outputs each block after encoding to the output unit 22.

  Further, the normal encoding processing unit 20 once decodes each block after encoding related to each intra frame, and stores the decoded image in the intra image frame memory unit 24.

  The inter-coding target image setting unit 28 has at least some pixel values of each pixel constituting the first frame, at least part of each pixel constituting each of the 2nd to (n-1) th frames, and n The pixel value is determined based on at least a part of the pixel values of each pixel constituting the th frame. Then, based on each determined pixel value, an inter-coding target image is generated. This will be specifically described below.

  The pixel value acquisition unit 280 acquires, from the intra image frame memory unit 24, the pixel value of each pixel constituting the first frame and the pixel value of each pixel constituting the nth frame, and the pixel value The data is output to the determination unit 281.

  The pixel value determination unit 281 sequentially pays attention to a plurality of blocks determined when the pixel value determination is performed by the processing target determination unit 18. Based on at least a part of each pixel constituting the target block and at least a part of each pixel constituting the part acquired by the corresponding part acquisition unit 16 in the first frame, the second to n− Among the first frames, the pixel values of at least a part of the pixels constituting each part acquired by the corresponding part acquisition unit 16 are determined.

  More specifically, the pixel value determining unit 281 sequentially acquires a plurality of pixels that constitute the target block one by one. Then, based on the portion corresponding to the target block for each of the 1st to (n-1) th frames acquired by the corresponding portion acquisition unit 16, each corresponding to each pixel in the target block acquired sequentially, Pixels in each of the 1st to (n-1) th frames are acquired.

  The pixel value determining unit 281 determines the pixel value of the pixel selected for the first frame, the pixel value of the pixel in the acquired target block, and the pixel value in the acquired target block for the pixels in the second to n−1th frames acquired in this way. The pixel value is determined based on the interpolation line formed by interpolating. Specifically, the pixel value determination unit 281 determines the pixel value based on the time interval between frames and the interpolation line.

  Acquisition of pixels in each of the first to (n-1) th frames will be described in detail with reference to FIGS. 2 and 3 again.

  In the example of FIG. 2, the pixel value determination unit 281 includes the portions P1a and P2a corresponding to the target block B as pixels in the first to n−1th frames corresponding to the pixel G4 included in the target block B. , P3a, the pixels G1a, G2a, G3a are obtained. The pixels G1a, G2a, and G3a are pixels that are in the same position as the position of the pixel G4 in the target block B in the portions P1a, P2a, and P3a, respectively.

  In the example of FIG. 3, the pixel value determination unit 281 includes the portions P1b and P2b corresponding to the target block B as pixels in the first to n−1th frames corresponding to the pixel G4 included in the target block B. , P3b, pixels G1b, G2b, G3b are obtained. The pixels G1b, G2b, and G3b are pixels that are in the same position as the position of the pixel G4 in the block of interest B in the portions P1b, P2b, and P3b, respectively. As a result, the positions of the pixels G1a, G2a, G3a in each frame and the positions of the pixels G1b, G2b, G3b in each frame are not necessarily the same position.

  Next, the interpolation will be described in detail with reference to FIGS.

  FIG. 4 shows an example of a pixel value of a certain pixel in the sixth frame and a pixel value of each pixel in each of the first to fifth frames corresponding to the pixel when n = 6. FIG. In the figure, one vertical line represents one pixel, and each pixel H1, H2, H3, H4, H5, H6 is the first, second, third, fourth, fifth, sixth, respectively. Pixels included in each frame. In the figure, the vertical line is indicated by a solid line or a dotted line, but the pixel indicated by the solid line is a pixel included in the intra frame, and the pixel indicated by the dotted line is a pixel included in the inter frame. Each pixel H1, H2, H3, H4, and H5 shown in the figure is acquired by the pixel value determining unit 281 as a pixel corresponding to the pixel H6.

  In addition, the luminance levels K1, K2, K3, K4, K5, and K6 in FIG. 4 indicate the pixel values of the pixels H1, H2, H3, H4, H5, and H6, respectively. As shown in the figure, the brightness levels K1, K2, K3, K4, K5, and K6 are slightly different.

  FIG. 5 illustrates a pixel H2 determined by the pixel value determination unit 281 based on an interpolation line obtained by interpolating the luminance level K1 of the pixel H1 and the luminance level K6 of the pixel H6 in the example illustrated in FIG. , H3, H4, and H5 luminance levels K2α, K3α, K4α, and K5α. In the figure, straight lines are used as interpolation lines. As shown in FIG. 5, the pixel value determination unit 281 sets the luminance level on the interpolation line as the luminance level of the pixels H2, H3, H4, and H5.

  FIG. 6 shows that the pixel value determination unit 281 in the example shown in FIG. 4 determines based on another interpolation line formed by interpolating the luminance level K1 of the pixel H1 and the luminance level K6 of the pixel H6. The luminance levels K2β, K3β, K4β, and K5β of the pixels H2, H3, H4, and H5 are shown. In the figure, a curve is used as an interpolation line.

  In the example of FIG. 6, the pixel value determination unit 281 acquires the luminance level (pixel value) of each of the second to n−1th pixels based on the image stored in the inter image frame memory unit 26. Then, based on the acquired luminance level (pixel value), an optimum interpolation line is determined by, for example, the least square method, and the luminance level (pixel value) on the determined interpolation line is changed from the second to the (n−1) th. It is determined as the luminance level (pixel value) of the pixel.

  As a result, as shown in FIG. 6, the luminance levels K2β, K3β, K4β, and K5β are closer to the original luminance levels K2, K3, K4, and K5 than the luminance levels K2α, K3α, K4α, and K5α shown in FIG. It is a value. Note that if the value is too close to the original pixel value, the effect of suppressing the occurrence of flicker is reduced, so it is necessary to select an optimal interpolation line as appropriate.

  The inter encoding target image setting unit 28 generates an inter encoding target image based on each pixel value determined as described above, and outputs the generated inter encoding target image to the inter image encoding unit 30.

  The inter image encoding unit 30 uses the input inter encoding target image (image composed of pixels having the pixel value determined by the pixel value determining unit 281) to each of the 2nd to n−1th frames. Is encoded. Specifically, for the block that does not include a portion corresponding to a block that is determined not to perform pixel value determination by the processing target determination unit 18 in the inter frame, the above-described predictive encoding, orthogonal transform, quantization, And an encoding process including each process of motion prediction. When the encoding process is completed, the inter image encoding unit 30 outputs the encoded blocks to the output unit 22.

  The output unit 22 acquires a series of encoded blocks based on the blocks respectively input from the normal encoding processing unit 20 and the inter image encoding unit 30 and sequentially outputs them. Each encoded block output in this way is stored in a storage medium (not shown) and decoded when it is reproduced.

  As described above, according to the encoding device 1, each of the 2nd to (n-1) th pixels is based on the pixel value of the pixel constituting the first frame and the pixel value of the pixel constituting the nth frame. Since the pixel values of the pixels constituting the frame can be determined, for example, the interval between them can be smoothly connected, and the occurrence of flicker can be suppressed in the nth frame that is an intra frame. In addition, since the flicker generation suppressing process does not require a process for changing the prediction mode of the nth frame, it is possible to suppress the deterioration of the intra-frame coding efficiency.

  Also, according to the encoding device 1, a smooth line between the pixel value of the pixel constituting the first frame and the pixel value of the pixel constituting the nth frame is obtained by an interpolation line which is a straight line or an arbitrary curve. It will be possible to connect to.

  Furthermore, according to the encoding device 1, the pixel for determining the pixel value can be appropriately determined based on the correspondence between the block of interest and each part.

  Further, according to the encoding device 1, the pixel values of the pixels constituting the second to (n-1) th frames are the pixel values of the pixels constituting the first frame and the pixels constituting the nth frame. Since the pixel value is determined based on the pixel value, in some cases, there is a possibility that the processing target determination unit 18 may be far from the original pixel value. Since the determination of whether or not the pixel value is determined by the pixel value determining unit 281 is determined based on the above, it is possible to prevent the pixel value determining unit 281 from determining the pixel value when they are not similar, for example. As a result, it is possible to prevent the pixel values of the pixels constituting the second to (n-1) th frames from being far from the original pixel values.

  Further, according to the encoding device 1, as the corresponding portion, a portion in the same position as the target block can be used in each frame. Alternatively, the corresponding part can be acquired based on the motion vector, and in this case, even if the image included in the block of interest moves over time in the frame, the corresponding part is acquired. The unit 16 can acquire an appropriate corresponding part.

  Furthermore, according to the encoding device 1, it is possible to suppress the occurrence of flicker for both luminance and color difference (red color difference and blue color difference) used in the color difference method.

The present invention is not limited to the above embodiment. For example, by recording a program for realizing the function of the encoding device 1 on a computer-readable recording medium, causing the computer system to read and execute the program recorded on the recording medium, the encoding device described above Each process of 1 may be performed.
Here, the “computer system” may include an OS and hardware such as peripheral devices. Further, the “computer system” includes a homepage providing environment (or display environment) if a WWW system is used.
The “computer-readable recording medium” means a flexible disk, a magneto-optical disk, a ROM, a writable nonvolatile memory such as a flash memory, a portable medium such as a CD-ROM, a hard disk built in a computer system, etc. This is a storage device.
Furthermore, the “computer-readable recording medium” includes a volatile memory (for example, DRAM (DRAM) in a computer system that becomes a server or a client when a program is transmitted through a network such as the Internet or a communication line such as a telephone line. Dynamic Random Access Memory)), etc., which hold programs for a certain period of time.
The program may be transmitted from a computer system storing the program in a storage device or the like to another computer system via a transmission medium or by a transmission wave in the transmission medium. Here, the “transmission medium” for transmitting the program refers to a medium having a function of transmitting information, such as a network (communication network) such as the Internet or a communication line (communication line) such as a telephone line.
Further, the program may be for realizing a part of the functions described above. Furthermore, what can implement | achieve each function mentioned above in combination with the program already recorded on the computer system, what is called a difference file (difference program) may be sufficient.

It is a schematic block diagram which shows the system configuration | structure and functional block of the encoding apparatus concerning embodiment of this invention. It is a figure which shows the specific example of the corresponding | compatible part acquired by the corresponding | compatible part acquisition part concerning embodiment of this invention. It is a figure which shows the specific example of the corresponding | compatible part acquired by the corresponding | compatible part acquisition part concerning embodiment of this invention. In the embodiment of the present invention, when n = 6, the pixel value of a pixel in the sixth frame, the pixel value of each pixel in each of the first to fifth frames corresponding to the pixel, It is a figure which shows the example of. It is a figure which shows the example of the interpolation by the pixel value determination part concerning embodiment of this invention. In the figure, straight lines are used as interpolation lines. It is a figure which shows the example of the interpolation by the pixel value determination part concerning embodiment of this invention. In the figure, a curve is used as an interpolation line.

Explanation of symbols

1 encoding device,
10 Intra / inter image determination unit,
12 Macroblock division part,
14 frame delay section,
16 Corresponding part acquisition unit,
18 processing object judgment part,
20 normal encoding processing unit,
22 output section,
24 Intra-image frame memory unit,
26 Inter image frame memory,
28 inter-coding target image setting unit,
30 Inter image encoding unit,
160 Intra macroblock similarity comparison unit,
161 Inter-macroblock similarity comparison unit,
280 pixel value acquisition unit,
281 Pixel value determination unit.

Claims (7)

Frame acquisition means for sequentially acquiring a series of frames, which are composed of n frames (n ≧ 3) constituting the moving image data, and the nth frame is an object of intra coding;
For at least some of the pixels constituting each of the 2nd to (n-1) th frames, at least some pixel values of each of the pixels constituting the first frame and at least each of the pixels constituting the nth frame A pixel value determining means for determining the pixel value based on the partial pixel value;
Encoding means for encoding each of the second to (n-1) th frames using the pixel value determined by the pixel value determining means;
An encoding device comprising: The encoding device according to claim 1, wherein
The pixel value determining means includes one pixel value of each pixel constituting the first frame and one of the pixels constituting the second to (n-1) th frames, and the nth Determining the pixel value based on an interpolation line formed by interpolating one pixel value of each pixel constituting the frame;
An encoding apparatus characterized by that. The encoding device according to claim 1 or 2,
Dividing means for dividing the nth frame into a plurality of blocks;
Corresponding part acquisition means for acquiring a part corresponding to the target block, which is one of the blocks, for each of the 1st to (n-1) th frames;
Including
The pixel value determining means includes at least a part of pixel values of each pixel constituting the block of interest and at least a part of each pixel constituting the part obtained by the corresponding part obtaining means in the first frame. And at least a part of pixel values constituting each part acquired by the corresponding part acquisition means among the second to (n-1) th frames, based on the pixel value of
An encoding apparatus characterized by that. The encoding device according to claim 3,
Similarity for calculating the similarity between the image content of the block of interest and the image content of each part acquired for at least part of each of the first to (n-1) -th frames by the corresponding part acquisition unit A calculation means;
Determining means for determining whether to perform pixel value determination by the pixel value determining means for the block of interest according to the similarity calculated by the similarity calculating means;
Including
The pixel value determining unit, when the determining unit determines to determine a pixel value for the block of interest, at least a part of each pixel constituting the block of interest and the first frame of the first frame Based on at least a part of each pixel constituting the part acquired by the corresponding part acquisition unit, each part acquired by the corresponding part acquisition unit among the second to (n-1) th frames is configured. Determining a pixel value of at least a portion of each pixel;
An encoding apparatus characterized by that. In the encoding device according to claim 3 or 4,
The corresponding part acquisition unit acquires, as a part corresponding to the target block, a part at the same position as the target block in the nth frame in each of the first to (n-1) th frames. To
An encoding apparatus characterized by that. In the encoding device according to claim 3 or 4,
The corresponding part acquisition means includes
Motion vector acquisition means for acquiring a motion vector indicating the position in each of the first to (n-1) th frames of the content of the image included in the block of interest;
A portion corresponding to the block of interest is obtained for each of the first to (n-1) th frames based on the motion vector.
An encoding apparatus characterized by that. Frame acquisition means for sequentially acquiring n (n ≧ 3) frames constituting the moving image data;
For at least some of the pixels constituting each of the 2nd to (n-1) th frames, at least some pixel values of each of the pixels constituting the first frame and at least each of the pixels constituting the nth frame Based on a part of the pixel values, pixel value determining means for determining the pixel values, and using the pixel values determined by the pixel value determining means, the second to n−1th frames are encoded Encoding means for
As a program to make the computer function.

Images