JP2014017621A - Video compression format converter, video compression format conversion method and program - Google Patents

Abstract

PROBLEM TO BE SOLVED: To convert a video compression format while suppressing degradation of image quality, even between compression systems where the unit block size of encoding is different before and after conversion.SOLUTION: The video compression format converter includes a prediction method candidate determination unit 212, a unit prediction value generation unit 213, a cost value calculation unit 214, and a prediction method determination unit 215. The prediction method candidate determination unit 212 selects one or more of all prediction methods applicable to a post conversion unit block as prediction method candidates for each post conversion unit block, i.e., a unit block of encoding after conversion. The unit prediction value generation unit 213 and cost value calculation unit 214 determine the application results of a prediction method candidate, for each post conversion unit block. The prediction method determination unit 215 determines a prediction method optimal to the post conversion unit block among the prediction method candidate, based on the results obtained from unit prediction value generation unit 213 and cost value calculation unit 214, for each post conversion unit block.

Description

  The present invention relates to a video compression format conversion device, a video compression format conversion method, and a program.

  Research and development on high-performance video compression technology is ongoing worldwide. New technologies are being standardized every few years as international standards for video compression technology.

  Recently, in 2003, H.C. H.264 (for example, see Non-Patent Document 1) has been standardized and is becoming widespread. However, in the field of international standardization, HEVC (for example, see Non-Patent Document 2) is being discussed as a new compression technique.

  H. Although a similar technique is used between different compression schemes such as H.264 and HEVC, there are cases where there is no compatibility as a format of compressed data. For example, H.C. Both H.264 and HEVC are based on motion compensation and orthogonal transform (discrete cosine transform), but there is no compatibility with respect to the format.

  For this reason, when a new video compression method becomes widespread, video contents having different compression data formats are mixed. Therefore, there is a need for a format conversion technique for converting video content from an old compressed data format to a new compressed data format.

Joint Video Team (JVT) of ISO / IEC MPEG and ITU-T VCEG, “Text of ISO / IEC 14496-10 Advanced Video Coding,” High Efficiency Video Coding (HEVC) text specification draft 6, JCTVC-H1003 ISO / IEC MPEG and ITU-T VCEG, “Text of ISO / IEC 13818-2 Generic coding of moving pictures and associated audio information: Video,” 2000.

  The conventional format conversion technique is a technique that assumes a case where the unit block size of the encoding process is the same before and after conversion.

  For example, MPEG-2 (for example, see Non-Patent Document 3) and H.264. In H.264, the unit block size of the encoding process is the same. For this reason, video content is changed from MPEG-2 to H.264. When converting to H.264, since the encoding processing blocks to be converted correspond one-to-one, it is easy to use the encoding information before conversion for the compression processing after conversion, and the conventional format conversion technology is applied. It was possible to do.

  However, H.C. When the unit block size of the encoding process is different before and after the conversion, such as H.264 to HEVC, if the encoded information before the conversion is simply used for the compression process after the conversion as described above, the image quality is deteriorated. There was a case.

  Therefore, the present invention has been made in view of the above-described problems, and a video compression format is suppressed while suppressing deterioration in image quality even between compression methods having different unit block sizes of encoding processing before and after conversion. An object is to provide a video compression format conversion device, a video compression format conversion method, and a program that can be converted.

The present invention proposes the following matters in order to solve the above problems.
(1) The present invention relates to a video compression format conversion device (for example, the video compression format conversion device 1 in FIG. 1 or the video compression format conversion device 1 in FIG. All of the prediction methods (e.g., FIG. 6 to FIG. 6) that can be applied to each unit block after conversion, which is a unit block for encoding processing after conversion. Prediction method candidate determination means (for example, prediction method candidate determination in FIG. 4) that selects one or more as prediction method candidates from nine Intra8 × 8, Intra16 × 16, Inter8 × 8, and Inter16 × 16) Unit 212 and the prediction method candidate determination unit 212A of FIG. 12) and the prediction method candidates for each unit block after conversion. Application result calculation means for obtaining an applied result (for example, equivalent to the unit predicted value generation unit 213 and the cost value calculation unit 214 in FIGS. 4 and 12) and the application result calculation means for each of the converted unit blocks. A prediction method determining means (for example, corresponding to the prediction method determining unit 215 in FIGS. 4 and 12) for determining an optimal prediction method for the converted unit block from the prediction method candidates based on the obtained results; A video compression format conversion device characterized by comprising:

  According to the present invention, a prediction method candidate determination unit, an application result calculation unit, and a prediction method determination unit are provided in a video compression format conversion device that converts a video compression format between compression methods having different unit block sizes for encoding processing. Was provided. Then, by the prediction method candidate determination means, for each unit block after conversion, which is a unit block of the encoding process after conversion, one or more prediction methods can be selected from all prediction methods applicable to the unit block after conversion It was decided to select as a candidate. Further, the result of applying the prediction method candidate is obtained for each unit block after conversion by the application result calculation means. Further, the prediction method determining means determines, for each unit block after conversion, the prediction method optimum for the converted unit block from prediction method candidates based on the result obtained by the application result calculation means. .

  Therefore, by selecting a prediction method that is expected to have high encoding performance in the encoding process that is the conversion process as a prediction method candidate, high encoding performance is expected in the encoding process for each unit block after conversion. Can be applied. Therefore, the video compression format can be converted while suppressing a decrease in image quality even between compression methods having different unit block sizes for encoding processing before and after conversion.

  Further, a result of applying the prediction method candidate is obtained by the application result calculation means, and based on this result, the prediction method optimum for the converted unit block is determined by the prediction method determination means. For this reason, the application result calculation means does not obtain application results for all prediction methods that can be applied in the converted unit block, but only for prediction method candidates that are a part of these prediction methods. . Therefore, it is possible to reduce the processing cost required for the conversion between the compression methods having different unit block sizes of the encoding process before and after the conversion.

  (2) In the video compression format conversion apparatus according to (1), the present invention includes all prediction methods that can be applied in the converted unit block, and a unit block before conversion that is a unit block of encoding processing before conversion. For each combination of all prediction methods applicable to the unit block after conversion (for example, equivalent to Intra4 × 4, Intra8 × 8, Inter4 × 4, and Inter8 × 8 in FIGS. 6 to 9) , Comprising a transition rate calculation means (e.g., equivalent to the prediction method candidate determination unit 212 in FIG. 4 or the prediction method candidate determination unit 212A in FIG. 12) that evaluates the transition of the prediction method before and after conversion by the transition rate, The prediction method candidate determination means proposes a video compression format conversion device characterized in that the prediction method candidate is selected based on the transition rate.

  Here, there may be a correlation between the prediction method in the unit block before conversion and the prediction method in the unit block after conversion.

  Therefore, according to the present invention, the transition rate calculation means is provided in the video compression format conversion apparatus of (1). Then, the transition rate calculation means applies all prediction methods that can be applied to the unit block after conversion and the one corresponding to the unit block after conversion among the unit blocks before conversion that is the unit block of the encoding process before conversion. For each combination of all prediction methods to be obtained, the transition of the prediction method before and after the conversion is evaluated by the transition rate. Further, the prediction method candidate determining means selects the prediction method candidate based on the transition rate.

  For this reason, the prediction method in the unit block before conversion can be reflected in the determination of the prediction method in the unit block after conversion. Therefore, it is possible to further suppress the deterioration of the image quality.

  (3) In the video compression format conversion apparatus according to (2), the present invention is an image analysis unit that performs, for each unit block after conversion, feature value analysis of a decoded video in a unit block before conversion corresponding to the unit block after conversion. (E.g., corresponding to the image analysis unit 211 in FIG. 4), and the transition rate calculation means calculates the transition rate based on the result of the feature amount analysis by the image analysis means. A format converter is proposed.

  Here, there is a correlation between the prediction method in the pre-conversion unit block and the prediction method in the post-conversion unit block according to the feature amount of the video, such as texture complexity and object tracking. There is.

  Therefore, according to the present invention, image analysis means is provided in the video compression format conversion apparatus of (2). Then, the feature amount analysis of the decoded video in the unit block before conversion corresponding to the unit block after conversion is performed for each unit block after conversion by the image analysis unit. Further, the transition rate is calculated by the transition rate calculation means based on the result of the feature amount analysis by the image analysis means.

  Therefore, the prediction method in the unit block before conversion can be reflected in the determination of the prediction method in the unit block after conversion according to the feature amount of the decoded video in the unit block before conversion corresponding to the unit block after conversion. Therefore, it is possible to further suppress the deterioration of the image quality.

  (4) In the video compression format conversion device according to (3), the image analysis unit performs the feature amount analysis, and the texture of the decoded video in the unit block before conversion corresponding to the unit block after conversion The transition rate calculating means calculates the transition rate based on the texture complexity evaluation result, and the prediction method candidate determining means determines the texture complexity based on the transition rate. A video compression format conversion device is proposed in which a prediction method that becomes more accurate as the value of the image becomes higher is selected as the prediction method candidate.

  According to the present invention, in the video compression format conversion apparatus of (3), the feature amount analysis is performed by the image analysis means, and the complexity of the texture is evaluated for the decoded video in the unit block before conversion corresponding to the unit block after conversion. The transition rate is calculated by the transition rate calculation means based on the texture complexity evaluation result. Further, the prediction method candidate determining means selects, as the prediction method candidate, a prediction method that becomes more accurate as the texture complexity increases based on the transition rate.

  Therefore, the prediction method in the unit block before conversion may be reflected in the determination of the prediction method in the unit block after conversion according to the complexity of the texture of the decoded video in the unit block before conversion corresponding to the unit block after conversion. it can. Therefore, it is possible to further suppress the deterioration of the image quality.

  (5) In the video compression format conversion apparatus according to (3) or (4), the image analysis unit performs the feature amount analysis and decodes the decoded video in the unit block before conversion corresponding to the unit block after conversion. The tracking rate of the object is evaluated, the transition rate calculating unit calculates the transition rate based on the evaluation result of the tracking property of the object, and the prediction method candidate determining unit is based on the transition rate, A video compression format conversion device has been proposed in which the degree of selecting intra prediction as the prediction method candidate is increased as the object followability decreases.

  According to the present invention, in the video compression format conversion device of (3) or (4), the feature amount analysis is performed by the image analysis unit, and the decoded video in the unit block before conversion corresponding to the unit block after conversion The followability is evaluated, and the transition rate is calculated by the transition rate calculation means based on the evaluation result of the object followability. Further, the prediction method candidate determining means increases the degree of selecting intra prediction as a prediction method candidate as the object followability decreases based on the transition rate.

  Therefore, the prediction method in the unit block before conversion may be reflected in the determination of the prediction method in the unit block after conversion according to the followability of the decoded video object in the unit block before conversion corresponding to the unit block after conversion. it can. Therefore, it is possible to further suppress the deterioration of the image quality.

  (6) In the video compression format conversion apparatus according to (1), the present invention provides the prediction method candidate determination unit with information indicating a prediction method to be selected as the prediction method candidate (for example, prediction in FIGS. 10 and 11). The prediction method candidate determination means proposes a video compression format conversion device that selects the prediction method candidate based on the information.

  According to the present invention, in the video compression format conversion device of (1), information indicating a prediction method to be selected as a prediction method candidate is input to the prediction method candidate determination unit, and the prediction method candidate determination unit performs the above-described operation. The prediction method candidate is selected based on the information.

  For this reason, an appropriate prediction method can be selected as a prediction method candidate based on the above information.

  (7) The present invention provides at least one of a prediction block size, an encoding method, an intra prediction direction, and a motion vector for the video compression format conversion device according to any one of (1) to (6). A video compression format conversion device is proposed which is used as the prediction method.

  According to the present invention, in the video compression format conversion device according to any one of (1) to (6), at least one of a prediction block size, an encoding method, an intra prediction direction, and a motion vector is predicted. We decided to use it as a method. According to this, an effect similar to the effect mentioned above can be produced.

  (8) The present invention provides a prediction method candidate determination unit (for example, equivalent to the prediction method candidate determination unit 212 in FIG. 4 or the prediction method candidate determination unit 212A in FIG. 12), an application result calculation unit (for example, FIG. 4 or FIG. 12 unit predicted value generation unit 213 and cost value calculation unit 214), and prediction method determination means (for example, equivalent to prediction method determination unit 215 in FIG. 4 and FIG. 12), and the unit block size of the encoding process Video compression format conversion in a video compression format conversion apparatus (for example, equivalent to the video compression format conversion apparatus 1 in FIG. 1 or the video compression format conversion apparatus 1A in FIG. 10) for converting a video compression format between different compression methods. In the method, the prediction method candidate deciding means performs the conversion for each unit block after conversion, which is a unit block of encoding processing after conversion. One or more of the prediction methods that can be applied in the rear unit block (for example, equivalent to Intra8 × 8, Intra16 × 16, Inter8 × 8, and Inter16 × 16 in FIGS. 6 to 9) as prediction method candidates The first step to select (e.g., corresponding to the prediction method candidate determination process of FIGS. 5 and 13) and the application result calculation means obtain the result of applying the prediction method candidate for each unit block after conversion. The second step (e.g., corresponding to the prediction value generation process and the cost value calculation process in FIGS. 5 and 13) and the prediction method determination means are obtained in the second step for each unit block after conversion. A third step of determining an optimal prediction method for the converted unit block from the prediction method candidates (for example, the prediction methods of FIGS. 5 and 13) And equivalent) in determination processing, it proposes a video compression format conversion method characterized by comprising a.

  According to this invention, one prediction method that can be applied to each unit block after conversion, which is a unit block for encoding processing after conversion, is selected from among all prediction methods that can be applied to this unit block after conversion. The above was selected as a prediction method candidate. Further, the result of applying the prediction method candidate is obtained for each unit block after conversion by the application result calculation means. Further, the prediction method determining means determines, for each unit block after conversion, the prediction method optimum for the converted unit block from prediction method candidates based on the result obtained by the application result calculation means. . For this reason, the effect similar to the effect mentioned above can be produced.

  (9) The present invention provides a prediction method candidate determination unit (for example, equivalent to the prediction method candidate determination unit 212 in FIG. 4 or the prediction method candidate determination unit 212A in FIG. 12), an application result calculation unit (for example, FIG. 4 or FIG. 12 unit predicted value generation unit 213 and cost value calculation unit 214), and prediction method determination means (for example, equivalent to prediction method determination unit 215 in FIG. 4 and FIG. 12), and the unit block size of the encoding process Video compression format conversion in a video compression format conversion apparatus (for example, equivalent to the video compression format conversion apparatus 1 in FIG. 1 or the video compression format conversion apparatus 1A in FIG. 10) for converting a video compression format between different compression methods. A program for causing a computer to execute the method, wherein the prediction method candidate determining means is a unit block of the encoding process after conversion. For each post-conversion unit block that is a block, all prediction methods applicable to the post-conversion unit block (e.g., equivalent to Intra 8 × 8, Intra 16 × 16, Inter 8 × 8, and Inter 16 × 16 in FIGS. 6 to 9) ) And a first step of selecting one or more as prediction method candidates (for example, equivalent to the prediction method candidate determination process in FIG. 5 or FIG. 13), and the application result calculation means includes the converted unit block A second step (for example, corresponding to the prediction value generation process and the cost value calculation process in FIGS. 5 and 13) for obtaining a result of applying the prediction method candidate, and the prediction method determination means, after the conversion For each unit block, the optimum prediction method for the converted unit block is determined from the prediction method candidates based on the result obtained in the second step. 3 step (e.g., corresponding to the prediction method determining process of FIG. 5 and FIG. 13) proposes a program for executing a, to the computer.

  According to this invention, by executing the program using a computer, the prediction method candidate determining means applies the converted unit block to each unit block after conversion, which is a unit block of the encoding process after conversion. One or more of the obtained prediction methods are selected as prediction method candidates. Further, the result of applying the prediction method candidate is obtained for each unit block after conversion by the application result calculation means. Further, the prediction method determining means determines, for each unit block after conversion, the prediction method optimum for the converted unit block from prediction method candidates based on the result obtained by the application result calculation means. . For this reason, the effect similar to the effect mentioned above can be produced.

  According to the present invention, it is possible to convert a video compression format while suppressing deterioration in image quality and reducing the processing cost required for conversion, even between compression methods having different encoding unit block sizes before and after conversion. it can.

1 is a block diagram of a video compression format conversion apparatus according to a first embodiment of the present invention. It is a block diagram of the pre-conversion decoding process part provided in the said video compression format converter. It is a block diagram of the conversion process part provided in the said video compression format converter. It is a block diagram of the 2nd predicted value generation part provided in the above-mentioned conversion processing part. It is a figure which shows the prediction process which the said 2nd predicted value production | generation part performs. It is a figure for demonstrating operation | movement of a said 2nd predicted value production | generation part. It is a figure for demonstrating operation | movement of a said 2nd predicted value production | generation part. It is a figure for demonstrating operation | movement of a said 2nd predicted value production | generation part. It is a figure for demonstrating operation | movement of a said 2nd predicted value production | generation part. It is a block diagram of the video compression format conversion apparatus which concerns on 2nd Embodiment of this invention. It is a block diagram of the conversion process part provided in the said video compression format converter. It is a block diagram of the 2nd predicted value generation part provided in the above-mentioned conversion processing part. It is a figure which shows the prediction process which the said 2nd predicted value production | generation part performs.

  Hereinafter, embodiments of the present invention will be described with reference to the drawings. Note that the constituent elements in the following embodiments can be appropriately replaced with existing constituent elements, and various variations including combinations with other existing constituent elements are possible. Accordingly, the description of the following embodiments does not limit the contents of the invention described in the claims.

<First Embodiment>
[Configuration and Operation of Video Compression Format Conversion Apparatus 1]
FIG. 1 is a block diagram of a video compression format conversion apparatus 1 according to the first embodiment of the present invention. The video compression format conversion apparatus 1 converts the pre-conversion compressed data a into post-conversion compressed data b of a compression method having a unit block size different from that of the pre-conversion compressed data a. The video compression format conversion apparatus 1 includes a pre-conversion decoding processing unit 10 and a conversion processing unit 20.

(Configuration and operation of pre-conversion decoding processor 10)
FIG. 2 is a block diagram of the pre-conversion decoding processing unit 10. The pre-transform decoding processing unit 10 includes an entropy decoding unit 11, a first predicted value generation unit 12, an inverse DCT / inverse quantization unit 13, and a local memory 14.

  The entropy decoding unit 11 receives the pre-conversion compressed data a as an input. The entropy decoding unit 11 performs entropy decoding on the pre-conversion compressed data a, extracts pre-conversion encoding information d and a residual signal f for each unit block before conversion, which is a unit block of encoding processing before conversion, Output.

  The first predicted value generation unit 12 receives pre-conversion encoded information d and a post-conversion decoded video e output from the local memory 14 to be described later. The first prediction value generation unit 12 generates and outputs a prediction value g from the pre-conversion decoded video e for each pre-conversion unit block according to the pre-conversion encoding information d.

  The inverse DCT / inverse quantization unit 13 receives the residual signal f. The inverse DCT / inverse quantization unit 13 performs inverse DCT processing and inverse quantization processing on the residual signal f for each unit block before conversion, and outputs the result as inverse DCT and inverse quantized residual signal h. To do.

  The predicted value g output from the first predicted value generation unit 12 and the residual signal h output from the inverse DCT / inverse quantization unit 13 are added by an adder for each unit block before conversion, and before conversion. The decoded video e is output from the pre-conversion decoding processing unit 10.

  The local memory 14 receives the decoded video e before conversion. The local memory 14 accumulates the input pre-conversion decoded video e and supplies it to the first predicted value generation unit 12 as appropriate.

(Configuration and operation of conversion processing unit 20)
FIG. 3 is a block diagram of the conversion processing unit 20. The transformation processing unit 20 includes a second predicted value generation unit 21, a DCT / quantization unit 22, an entropy encoding unit 23, an inverse DCT / inverse quantization unit 24, and a local memory 25.

  The second predicted value generation unit 21 includes a post-conversion local decoded video n output from the local memory 25, an encoding parameter c input from outside the conversion processing unit 20, and pre-conversion encoding information d. , And the decoded video e before conversion. As shown in FIG. 4, the second prediction value generation unit 21 includes an image analysis unit 211, a prediction method candidate determination unit 212, a unit prediction value generation unit 213, a cost value calculation unit 214, and a prediction method determination unit 215. 5 for each unit block after conversion, which is a unit block for encoding processing after conversion, and the optimal prediction method for the current unit block after conversion is determined and determined. A prediction value i by the prediction method and prediction information j indicating the determined prediction method are output.

  In this embodiment, examples of the prediction method include a prediction block size, a coding method, a prediction direction in intra coding, a motion vector, and the like. Examples of the encoding method include intra prediction and inter prediction.

  The image analysis unit 211 performs the image analysis process of FIG. The image analysis unit 211 receives the pre-conversion decoded video e as an input, and performs feature amount analysis of the decoded video of the unit block before conversion corresponding to the unit block after conversion for each unit block after conversion. In the feature amount analysis, at least one of a complexity of a texture to be described later and a followability of an object to be described later is evaluated, and an image feature amount is determined based on the evaluation result.

  Here, the pre-conversion unit block corresponding to the above-described post-conversion unit block will be described. For example, one of all the converted unit blocks constituting the frame is referred to as a predetermined converted unit block. Then, the pre-conversion unit block corresponding to the predetermined post-conversion unit block is a pre-conversion unit block including pixels included in the predetermined post-conversion unit block among all the pre-conversion unit blocks constituting the frame. is there.

  The evaluation of the complexity of the texture is suitable for evaluating the ease of intra prediction. When evaluating the complexity of the texture, for example, the variance of pixel values in the unit block before conversion and the result of applying a Laplacian filter to each pixel in the unit block before conversion are squared in the unit block before conversion. The energy of the AC component when the sum value and the pixel value in the unit block before conversion are frequency-converted by Fourier transform or the like is obtained.

  On the other hand, evaluation of object followability is suitable for evaluating the ease of inter prediction. When evaluating the object followability, first, the movement of the object is detected. In detecting the motion of an object, for example, the size of a motion vector when block matching is performed with a local decoded image of a past frame in time, or the size of a global motion in an area (or the entire frame) including a unit block before conversion. I ask for it. Note that in detection of motion vectors by block matching with temporally past frames, motion vector information included in pre-conversion encoded information that is an input to a prediction method candidate determination unit 212 described later can be used.

  Next, the sum of squares or the sum of absolute values related to the difference between the encoding processing block pixel value and the reference pixel value when the motion compensation of the encoding processing block is performed using the detected object motion is obtained.

  The prediction method candidate determination unit 212 performs the prediction method candidate determination process of FIG. The prediction method candidate determination unit 212 receives as input the pre-conversion encoding information d, the encoding parameter c, and the image feature amount determined by the above-described image analysis processing. For each unit block after conversion, the prediction method candidate determination unit 212 first extracts a prediction method effective in the unit block after conversion based on the pre-conversion encoding information d. Next, a prediction method finally set as a candidate in the unit block after conversion is determined as prediction method candidate information using the image feature amount and the encoding parameter c.

  Here, the video compression format conversion apparatus 1 according to the present embodiment assumes a case where the unit block size of the encoding process is different before and after conversion of the video compression format. Therefore, the prediction method candidate determination unit 212 encodes all the prediction methods in the unit block before conversion corresponding to the unit block after conversion as a prediction method candidate suitable for the unit block after conversion, and each of these prediction methods. Select a prediction method with a similar performance.

  Specifically, the prediction method candidate determination unit 212 first has all the prediction methods that can be applied in the unit block after conversion, and all the prediction methods that can be applied in the unit block before conversion corresponding to the unit block after conversion. For each combination, the transition rate of the prediction method before and after conversion is set to an initial value, and a threshold value for these transition rates is set.

  FIG. 6 is a diagram illustrating an example in which transition rates are set to initial values and threshold values for these transition rates are set. In FIG. 6, Intra4 × 4, Intra8 × 8, Inter4 × 4, and Inter8 × 8 are assumed as the prediction methods before conversion, and the prediction methods after conversion are Intra8 × 8, Intra16 × 16, Inter8 × 8, And Inter16 × 16 are assumed. For example, 15% is set as the initial value of the transition rate for transition from Intra4 × 4 to Intra8 × 8, and 5% is set as the initial value of the transition rate for transition from Inter4 × 4 to Inter16 × 16. Further, 10% is set as the threshold of the transition rate.

  Next, the prediction method candidate determination unit 212 updates the transition rate based on the image feature amount determined by the image analysis unit 211 and the encoding parameter c.

  FIG. 7 is a diagram illustrating an example in which the transition rate illustrated in FIG. 6 is updated using the complexity of the texture. In FIG. 7, as the texture complexity increases, the transition rate of the prediction method for subdividing the prediction block is increased. According to this, as the texture complexity increases, a prediction method that becomes more accurate is selected as a prediction method candidate by the prediction method candidate determination unit 212 described later.

  FIG. 8 is a diagram showing an example in which the transition rate shown in FIG. 7 is updated by using the object followability. In FIG. 8, the transition rate of intra prediction is increased as the object followability decreases. According to this, as the object followability becomes lower, intra prediction is selected as a prediction method candidate in the prediction method candidate determination unit 212 described later.

  Next, the prediction method candidate determination unit 212 determines a prediction method having a transition rate equal to or higher than the threshold as a prediction method candidate in the converted processing block.

  FIG. 9 is a diagram illustrating an example in which a prediction method candidate in the converted processing block is selected based on the transition rate illustrated in FIG. In FIG. 9, there are five transition rates that are equal to or greater than the threshold value (10%) set above. For this reason, Intra8 × 8, Intra16 × 16, and Inter8 × 8, which are determined to transition at these five transition rates, are selected as prediction method candidates.

  The unit predicted value generation unit 213 performs the predicted value generation process of FIG. The unit predicted value generation unit 213 receives the pre-conversion decoded video e, the post-conversion local decoded video n, and the prediction method candidate determined by the prediction method candidate determination process described above as input, and for each prediction method candidate, According to the prediction method, a predicted value is obtained from the input video.

  The cost value calculation unit 214 performs the cost value calculation process of FIG. The cost value calculation unit 214 receives the prediction value for each prediction method candidate obtained in the prediction value generation process, the converted local decoded video n, and the pre-conversion decoded video e, and inputs each prediction method candidate. The coding distortion and code amount generated by the prediction method are calculated as cost values.

  The prediction method determination unit 215 performs the prediction method determination process of FIG. The prediction method determination unit 215 receives the cost value for each prediction method candidate obtained in the cost value calculation process, the converted local decoded video n, and the pre-converted decoded video e, and inputs the prediction with the highest cost value. The prediction method of the method candidate is determined as the optimal prediction method for the current converted unit block, and the prediction value i by the determined prediction method and the prediction information j indicating the determined prediction method are output.

  Returning to FIG. 3, the DCT / quantization unit 22 receives a difference signal (residual signal) between the predicted value i and the pre-conversion decoded video e. The DCT / quantization unit 22 performs DCT on the residual signal for each unit block after conversion to obtain DCT coefficients, performs quantization processing on the DCT coefficients, and performs DCT and the quantized residuals. Output as a difference signal k.

  The entropy encoding unit 23 receives the prediction information j, the DCT, and the quantized residual signal k as inputs. The entropy encoding unit 23 performs variable length encoding or arithmetic encoding on the input information for each unit block after conversion, writes the result as a compressed data stream according to the encoding syntax, and after conversion Output as compressed data b.

  The inverse DCT / inverse quantization unit 24 receives the DCT and the quantized residual signal k as inputs. The inverse DCT / inverse quantization unit 24 performs inverse quantization processing and inverse DCT processing on the DCT and the quantized residual signal k for each unit block after conversion, and performs inverse quantization and inverse DCT. Output as a residual signal m.

  The local memory 25 receives the converted local decoded video n. The post-conversion local decoded video n is addition information of the predicted value i and the residual signal m that has been subjected to inverse quantization and inverse DCT. The local memory 25 accumulates the input converted local decoded video n and supplies it to the second predicted value generation unit 21 as appropriate.

  According to the video compression format converter 1 described above, the following effects can be obtained.

  For each unit block after conversion, the video compression format conversion device 1 evaluates at least one of texture complexity and object followability for the decoded video of the unit block before conversion corresponding to the unit block after conversion. Thus, the image feature amount is determined based on the evaluation result. Then, out of all the prediction methods that can be applied in the unit block after conversion, the candidates in the unit block after conversion are narrowed down using the image feature amount, a cost value is calculated for each candidate, and based on the cost value Then, one of the narrowed candidates is determined as a prediction method to be applied in the unit block after conversion. Therefore, in determining the prediction method in the unit block after conversion, the prediction method in the unit block before conversion is determined according to the complexity of the texture of the decoded video in the unit block before conversion corresponding to the unit block after conversion and the followability of the object. It can be reflected. Therefore, it is possible to convert the video compression format while suppressing deterioration in image quality.

  In addition, as described above, the video compression format conversion apparatus 1 finally narrows down the prediction method to be applied to the converted unit block after narrowing down candidates from all the prediction methods that can be applied to the converted unit block. One is determined. For this reason, since the process which calculates a cost value can be decreased compared with the case where it does not narrow down as mentioned above, the process cost required for conversion can be reduced.

Second Embodiment
[Configuration and Operation of Video Compression Format Conversion Apparatus 1A]
FIG. 10 is a block diagram of a video compression format conversion apparatus 1A according to the second embodiment of the present invention. The video compression format conversion device 1A differs from the video compression format conversion device 1 according to the first embodiment of the present invention shown in FIG. 1 in that a conversion processing unit 20A is provided instead of the conversion processing unit 20. In the video compression format conversion device 1A, the same components as those of the video compression format conversion device 1 are denoted by the same reference numerals and description thereof is omitted.

(Configuration and operation of conversion processing unit 20)
FIG. 11 is a block diagram of the conversion processing unit 20A. The conversion processing unit 20A is different from the conversion processing unit 20 according to the first embodiment of the present invention illustrated in FIG. 3 in that the conversion processing unit 20A includes a second predicted value generation unit 21A instead of the second predicted value generation unit 21. Different.

  FIG. 12 is a block diagram of the second predicted value generation unit 21A. The second predicted value generation unit 21A is different from the second predicted value generation unit 21 according to the first embodiment of the present invention shown in FIG. The difference is that a prediction method candidate determination unit 212A is provided instead of the unit 212.

  The prediction method candidate determination unit 212A receives the pre-transform coding information d and the prediction method candidate information p as inputs. The prediction method candidate information p is information indicating a prediction method to be selected as a prediction method candidate, and is input from the outside of the conversion processing unit 20A. In the present embodiment, the transition of the prediction method before and after the conversion does not depend on the image feature amount.

  This prediction method candidate determination unit 212A selects, as prediction method candidate information, a prediction method that is finally made a candidate in a processing block according to the prediction method candidate information p for the prediction method of the pre-transform coding information d. . When selecting as prediction method candidate information, for example, all the prediction methods for all the prediction methods before conversion included in the processing block are candidates, or among the prediction methods before conversion included in the processing block, the frequency is high. A prediction method corresponding to the prediction method is set as a candidate.

  According to the video compression format conversion device 1A described above, the following effects can be obtained.

  Based on the prediction method candidate information p, the video compression format conversion device 1A narrows down the candidates in the converted unit block using image feature amounts from all the prediction methods that can be applied in the converted unit block. A cost value is calculated for each candidate, and based on the cost value, one of the narrowed candidates is determined as a prediction method to be applied in the converted unit block. For this reason, the same effect as the above-mentioned effect which the video compression format conversion apparatus 1 can show | play can be show | played.

  Further, in the video compression format conversion device 1A, the image analysis unit 211 provided in the video compression format conversion device 1 is unnecessary. For this reason, compared with the video compression format conversion apparatus 1, while being able to simplify a structure, the processing cost required for conversion can be further proposed.

  The processing of the video compression format conversion device 1 and the video compression format conversion device 1A of the present invention is recorded on a computer-readable non-transitory recording medium, and the program recorded on the recording medium is recorded on the video compression format conversion device. 1 and the video compression format conversion apparatus 1A are read and executed.

  Here, for example, a nonvolatile memory such as an EPROM or a flash memory, a magnetic disk such as a hard disk, a CD-ROM, or the like can be applied to the above-described recording medium. Further, reading and execution of the program recorded on the recording medium is performed by a processor provided in the video compression format conversion device 1 or the video compression format conversion device 1A.

  The above-mentioned program is transmitted from the video compression format conversion device 1 or the video compression format conversion device 1A 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. May be transmitted. 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 above-described program may be for realizing a part of the above-described function. Furthermore, what can implement | achieve the above-mentioned function in combination with the program already recorded on the video compression format converter 1 or the video compression format converter 1A, what is called a difference file (difference program) may be sufficient.

  The embodiment of the present invention has been described in detail with reference to the drawings. However, the specific configuration is not limited to this embodiment, and includes a design that does not depart from the gist of the present invention.

  For example, in the first embodiment and the second embodiment described above, it is assumed that the unit block size of the encoding process is different before and after the conversion of the video compression format. The processing block size may be smaller than the converted processing block size, or may be larger than the converted processing block size.

  In the first embodiment and the second embodiment described above, Intra4 × 4, Intra8 × 8, Inter4 × 4, and Inter8 × 8 are assumed as prediction methods before conversion, but are not limited thereto. Various prediction methods can be assumed. In addition, as prediction methods after conversion, Intra8 × 8, Intra16 × 16, Inter8 × 8, and Inter16 × 16 are assumed, but not limited thereto, various prediction methods can be assumed. .

  In the second embodiment described above, the prediction method candidate information p is input from the outside of the conversion processing unit 20A. However, the prediction method candidate information p is not limited thereto, and may be generated inside the conversion processing unit 20A. Good.

  In the first embodiment and the second embodiment described above, the compression method of the pre-conversion compressed data a is, for example, H.264. H.264, and the compression method of the converted compressed data b may be HEVC, for example. Further, the compression method of the pre-conversion compressed data a may be MPEG-2, for example, and the compression method of the post-conversion compressed data b may be HEVC, for example.

DESCRIPTION OF SYMBOLS 1, 1A ... Video compression format conversion apparatus 10 ... Pre-conversion decoding process part 20, 20A ... Conversion process part 21, 21A ... 2nd prediction value generation part 211 ... Image analysis part 212 212A ... prediction method candidate determination unit 213 ... unit prediction value generation unit 214 ... cost value calculation unit 215 ... prediction method determination unit

Claims (9)

A video compression format converter for converting a video compression format between compression methods having different unit block sizes for encoding processing,
Prediction method candidate determination means for selecting one or more prediction methods as prediction method candidates from all prediction methods applicable to the converted unit block for each unit block after conversion, which is a unit block for encoding processing after conversion. When,
Application result calculation means for obtaining a result of applying the prediction method candidate for each unit block after conversion;
For each converted unit block, a prediction method determining unit that determines an optimal prediction method for the converted unit block from the prediction method candidates based on the result obtained by the application result calculating unit. A video compression format converter characterized by the above. All prediction methods that can be applied in the unit block after conversion, and all prediction methods that can be applied in the unit block before conversion that is a unit block of encoding processing before conversion, corresponding to the unit block after conversion, For each combination of and, there is provided a transition rate calculation means for evaluating the transition of the prediction method before and after conversion by the transition rate,
The video compression format conversion apparatus according to claim 1, wherein the prediction method candidate determination unit selects the prediction method candidate based on the transition rate. For each unit block after conversion, comprising image analysis means for analyzing the feature amount of the decoded video in the unit block before conversion corresponding to the unit block after conversion,
3. The video compression format conversion apparatus according to claim 2, wherein the transition rate calculation unit calculates the transition rate based on a result of feature amount analysis by the image analysis unit. The image analysis means performs the feature amount analysis, evaluates the complexity of the texture for the decoded video in the unit block before conversion corresponding to the unit block after conversion,
The transition rate calculating means calculates the transition rate based on the evaluation result of the complexity of the texture,
The said prediction method candidate determination means selects the prediction method which becomes high precision as the complexity of the said texture becomes high based on the said transition rate as said prediction method candidate. Video compression format converter. The image analysis means performs the feature amount analysis, and evaluates the followability of the object for the decoded video in the unit block before conversion corresponding to the unit block after conversion,
The transition rate calculating means calculates the transition rate based on an evaluation result of the followability of the object,
5. The prediction method candidate determining unit increases the degree of selecting intra prediction as the prediction method candidate as the followability of the object decreases based on the transition rate. 6. Video compression format converter. Information indicating a prediction method to be selected as the prediction method candidate is input to the prediction method candidate determination unit,
The video compression format conversion apparatus according to claim 1, wherein the prediction method candidate determination unit selects the prediction method candidate based on the information.   7. The video compression format conversion according to claim 1, wherein at least one of a prediction block size, an encoding method, an intra prediction direction, and a motion vector is used as the prediction method. apparatus. Video compression format conversion in a video compression format conversion apparatus that includes a prediction method candidate determination unit, an application result calculation unit, and a prediction method determination unit, and performs conversion of a video compression format between compression methods having different unit block sizes of encoding processing A method,
The prediction method candidate determination means selects one or more prediction method candidates from all prediction methods applicable to the converted unit block for each unit block after conversion, which is a unit block for encoding processing after conversion. A first step of selecting as
A second step in which the application result calculation means obtains a result of applying the prediction method candidate for each unit block after conversion;
The prediction method determining means determines, for each unit block after conversion, an optimal prediction method for the unit block after conversion from the prediction method candidates based on the result obtained in the second step. And a video compression format conversion method comprising the steps of: Video compression format conversion in a video compression format conversion apparatus that includes a prediction method candidate determination unit, an application result calculation unit, and a prediction method determination unit, and performs conversion of a video compression format between compression methods having different unit block sizes of encoding processing A program for causing a computer to execute the method,
The prediction method candidate determination means selects one or more prediction method candidates from all prediction methods applicable to the converted unit block for each unit block after conversion, which is a unit block for encoding processing after conversion. A first step of selecting as
A second step in which the application result calculation means obtains a result of applying the prediction method candidate for each unit block after conversion;
The prediction method determining means determines, for each unit block after conversion, an optimal prediction method for the unit block after conversion from the prediction method candidates based on the result obtained in the second step. A program for causing a computer to execute step 3.

Images