JP5509048B2 - Intra prediction apparatus, encoder, decoder, and program - Google Patents

Description

  The present invention predicts a pixel value to be encoded from decoded pixel values by extrapolation, and in particular, an intra-prediction device, an encoder, and a decoding capable of prediction without impairing the continuity of the AC component of an image Device and program.

  MPEG-4 AVC / H. In the H.264 system (ISO / IEC 14496-10 / ITU-T Rec. H.264), in addition to inter prediction (inter-screen prediction), the pixel value of an adjacent block that has been encoded is determined for an image block to be encoded. Intra-prediction coding (intra-screen prediction coding) that generates a prediction image and encodes a difference from the prediction image is employed.

  FIG. 9 shows a conventional MPEG-4 AVC / H. The block diagram of the encoder 4 in a H.264 system is shown. The encoder 4 includes a subtraction unit 31, an orthogonal transformation unit 32, a quantization unit 33, a variable length coding unit 34, an inverse quantization unit 35, an inverse orthogonal transformation unit 36, a changeover switch 37, An intra prediction unit 38, a frame memory 39, a motion compensation prediction unit 40, and an addition unit 41 are provided.

  The motion compensation prediction unit 40 performs motion vector detection on the input image using the reference image acquired from the frame memory 39, performs motion compensation using the obtained motion vector, and obtains the predicted image obtained as a result. Is output to the subtracting unit 31 and the adding unit 41 via the changeover switch 37. In addition, the motion compensation prediction unit 40 outputs motion vector information to the variable length encoding unit 34.

  The subtraction unit 31 generates a difference image between the input image and the prediction image from the motion compensation prediction unit 40 or the intra prediction unit 38 and outputs the difference image to the orthogonal transformation unit 32.

  The orthogonal transform unit 32 performs orthogonal transform (for example, DCT; Discrete Cosine Transform) for each pixel block in the small region on the difference image supplied from the subtraction unit 31 and outputs the orthogonal transform coefficient to the quantization unit 33. .

  The quantization unit 33 selects a quantization table for the orthogonal transform coefficient input from the orthogonal transform unit 32, performs quantization processing, and outputs the quantization table to the variable length coding unit 34 and the inverse quantization unit 35.

  The variable length encoding unit 34 scans the quantized orthogonal transform coefficient input from the quantization unit 33 and performs variable length encoding processing to generate a bitstream, and also receives the input from the motion compensation prediction unit 41. The motion vector information is also subjected to variable length coding and output to the outside.

  The inverse quantization unit 35 performs an inverse quantization process on the quantized orthogonal transform coefficient input from the quantization unit 33 and outputs the result to the inverse orthogonal transform unit 36.

  The inverse orthogonal transform unit 36 performs inverse orthogonal transform (for example, IDCT; Inverse Discrete Cosine Transform) on the orthogonal transform coefficient input from the inverse quantization unit 35 and outputs the result to the addition unit 41.

  The addition unit 41 adds the image obtained by the inverse orthogonal transformation obtained from the inverse orthogonal transformation unit 36 and the prediction image obtained from the motion compensation prediction unit 40 or the intra prediction unit 38 to generate a decoded image, and generates an intra prediction unit. 38 and the frame memory 39.

  The changeover switch 37 switches between motion compensation prediction and intra prediction.

  The intra prediction unit 38 generates a prediction image predicted intra from the image obtained by decoding the already-encoded block (the output image of the addition unit 41), and outputs the prediction image to the subtraction unit 31 and the addition unit 41. Here, the subtraction unit 31 outputs the difference image between the predicted image and the original image to the orthogonal transform unit 32 and encodes it via the quantization unit 33 and the variable length coding unit 34.

  Intra prediction is performed in units of 4 pixels × 4 lines, 8 pixels × 8 lines, or 16 pixels × 16 lines, and a plurality of types of prediction modes (prediction directions) (for example, prediction in units of 4 pixels × 4 lines). Selects the optimum prediction direction from among nine types. FIG. 10 is a diagram illustrating a prediction mode when prediction is performed in units of 4 pixels × 4 lines. In the figure, hatched circles indicate decoded pixels, white circles indicate pixels to be predicted, and arrows indicate prediction directions. (A) Prediction mode 0 is vertical direction prediction, (b) Prediction mode 1 is horizontal direction prediction, (c) Prediction mode 2 is DC prediction, (d) Prediction mode 3 is diagonal lower left direction prediction, ( e) Prediction mode 4 in diagonal lower right direction prediction, (f) prediction mode 5 in vertical right direction prediction, (g) prediction mode 6 in horizontal down direction prediction, and (h) prediction mode 7 in vertical left direction. In the prediction mode 8 of (i), horizontal upward prediction is performed. The above is MPEG-4 AVC / H. This is a technique of intra prediction in the H.264 system.

  In addition, in order to increase the accuracy of intra prediction, a technique of encoding a difference value between adjacent pixels by always referring to adjacent pixels instead of referring to spatially separated pixels (for example, Patent Document 1). And a technique for performing intra prediction based on the value of a pixel adjacent to a prediction block and the value of a pixel between the prediction block and one or more pixels is known (see, for example, Patent Document 2). Also known is a technique for generating a value that takes into account the frequency characteristics of adjacent decoded images as a prediction value in intra coding in order to improve the prediction performance for a picture image whose pixel value has a certain change tendency. (For example, see Patent Document 3).

JP 2009-049969 A JP 2008-271371 A JP 2008-245088 A

  However, the conventional intra prediction method with directionality assumes that the value of the decoded adjacent pixel and the value of the prediction target pixel are continuous in a DC sense, and in an AC sense. Continuity was not assumed. Further, although the technique described in Patent Document 3 performs prediction including an AC component, there is no AC (phase) continuity between the reference pixel and the prediction pixel. For this reason, when the reference image is a part of a periodic pattern, it has not been possible to perform advanced prediction considering the AC component. In addition, there is a problem that the coding distortion included in the decoded image and the noise component included in the original image directly affect the predicted value.

  In order to solve the above-described problem, an object of the present invention is to provide an intra-prediction device that performs prediction that does not impair the continuity of the AC component between the reference pixel and the prediction pixel, and performs prediction with less influence of the noise component, and encoding A decoder, a decoder, and a program.

  In order to solve the above problems, an intra prediction apparatus according to the present invention is an intra prediction apparatus that generates a prediction image of a prediction block from an image of a decoded reference block adjacent to the prediction block, and an initial value of the prediction image A prediction image initial value setting unit that sets a prediction block, a processing block generation unit that generates a processing block including a reference block and a prediction block in which the initial value is set, and an orthogonal transform on the image of the processing block An orthogonal transform unit that generates an orthogonal transform coefficient, a coefficient correction unit that corrects the absolute value of the orthogonal transform coefficient generated by the orthogonal transform unit to a value smaller than the original value in a high frequency region, and the correction. Among the images generated by the inverse orthogonal transform unit, an inverse orthogonal transform unit that performs inverse orthogonal transform on the orthogonal transform coefficient and generates an image, An image correction unit that corrects an image area in a reference block to a pixel value of an original decoded image, supplies the corrected image to the orthogonal transform unit, and a prediction block from the image generated by the inverse orthogonal transform unit A predetermined condition for terminating the repetition of a series of processes by the predicted image extracting unit that extracts the image and generating a predicted image, and the image correcting unit, the orthogonal transform unit, the coefficient correcting unit, and the inverse orthogonal transform unit When it is determined that the predetermined condition is not satisfied, the image generated by the inverse orthogonal transform unit is repeatedly sent to the image correction unit to repeatedly execute the series of processes. A branching unit that terminates the series of processes and supplies the image generated by the inverse orthogonal transform unit to the predicted image extraction unit when it is determined that the predetermined condition is satisfied. , A reference image analysis unit that analyzes the frequency of the image of the reference block, and an area in which the coefficient correction unit corrects an orthogonal transform coefficient based on an analysis result obtained by the reference image analysis unit, or the image correction unit And a control unit that determines at least one of the regions for correcting the pixel value.

  Furthermore, in the intra prediction apparatus according to the present invention, the reference image analysis unit derives power (power) of a high-frequency component of the reference block, and the control unit reduces the coefficient correction unit as the derived power is smaller. Is determined such that at least one of a high-frequency region for correcting the orthogonal transform coefficient or an image region for correcting the pixel value by the image correction unit is increased.

  Further, in the intra prediction apparatus according to the present invention, the reference image analysis unit derives a horizontal high-frequency component power and a vertical high-frequency component power of the reference block, and the control unit derives the derived high-frequency component power. The smaller the power of the high-frequency component in the horizontal direction, the at least the horizontal width of the high-frequency region in which the coefficient correction unit corrects the orthogonal transform coefficient, or the horizontal width of the image region in which the image correction unit corrects the pixel value. The length of the high-frequency region in which the coefficient correction unit corrects the orthogonal transform coefficient is determined as the power of the derived vertical high-frequency component is smaller, or the image correction unit Is determined such that at least one of the lengths in the vertical direction of the image region whose pixel value is to be corrected becomes longer.

  Furthermore, in the intra prediction device according to the present invention, the coefficient correction unit has an absolute value of the orthogonal transform coefficient generated by the orthogonal transform unit smaller than an original value in a high frequency region including the highest order orthogonal transform coefficient. Each time it is corrected to a value and repeated, the area of the high-frequency region where the orthogonal transform coefficient is corrected is made equal to or less than the area before the repetition.

  Further, in the intra prediction device according to the present invention, the image correction unit corrects an image area including an image adjacent to the prediction block in the reference block to a pixel value of an original decoded image, Each time it is repeatedly executed, the area of the image region whose pixel value is corrected is set to be equal to or less than the area before the repetition.

  Furthermore, in the intra prediction apparatus according to the present invention, the predetermined condition for ending the series of processing is that the number of times of the series of processing is equal to or greater than the predetermined number of times, or an image of the reference block after the series of processing and the sequence. The difference in pixel value from the image of the reference block before performing the process is not more than a predetermined threshold value.

  Also, an encoder according to the present invention includes the intra prediction device described above.

  A decoder according to the present invention includes the above-described intra prediction device.

  In addition, the present invention provides a computer functioning as an intra prediction device that generates a predicted image of a predicted block from an image of a decoded reference block adjacent to the predicted block. (A) Analyzing the frequency of the image of the reference block (B) setting an initial value of a predicted image in a prediction block; (c) generating a processing block including the reference block and a prediction block in which the initial value is set; and (d) the processing. A step of performing orthogonal transformation on the image of the block to generate an orthogonal transformation coefficient; and (e) a value smaller than the original value in the high frequency region, the absolute value of the orthogonal transformation coefficient generated by the step (d). And (f) generating an image by performing inverse orthogonal transform on the orthogonal transform coefficient modified in the step (e). A step, (g) determining whether or not a predetermined condition for ending the repetition of a series of processes from step (d) to step (f) is satisfied, and (h) the step (g) When it is determined that the predetermined condition is not satisfied, the image area in the reference block is corrected to the pixel value of the original decoded image among the images generated in the step (f), and the step (D) repeatedly performing the step (g), and (i) if it is determined in step (g) that the predetermined condition is satisfied, the repetition of the series of processes is terminated. Extracting a predicted block image from the image generated in step (f) to generate a predicted image; and (j) based on the analysis result obtained in step (a). And a step of determining at least one of a high frequency region for correcting the orthogonal transformation coefficient in step (e) and an image region for correcting the pixel value in step (h). It is done.

  According to the present invention, it is possible to perform prediction without impairing the continuity of the AC component between the reference pixel and the prediction pixel. In addition, it is possible to perform prediction with little influence of coding distortion included in a decoded image and noise components included in an original image.

  In addition, by setting parameters based on the analysis result of the frequency characteristics of the reference image, an appropriate predicted image can be generated according to the frequency characteristics of the reference image.

It is a block diagram which shows the structure of the intra prediction apparatus by the side of the encoding of Example 1 by this invention. It is a figure explaining the process of the intra estimation apparatus of Example 1 by this invention. It is a block diagram which shows the structure which concerns on the control based on the reference image analysis of the intra prediction apparatus of Example 1 by this invention. It is a block diagram which shows the structure of the intra prediction apparatus by the side of the decoding of Example 1 by this invention. It is a flowchart which shows operation | movement of the intra prediction apparatus of Example 1 by this invention. It is a block diagram which shows the structure of an encoder provided with the intra prediction apparatus of Example 1 by this invention. It is a block diagram which shows the structure of a decoder provided with the intra prediction apparatus of Example 1 by this invention. It is a block diagram which shows the structure concerning the control based on the reference image analysis of the intra prediction apparatus of Example 2 by this invention. Conventional MPEG4 AVC / H. 2 is a block diagram illustrating a configuration of an encoder in the H.264 scheme. FIG. Conventional MPEG4 AVC / H. It is a figure which shows the prediction mode in the case of estimating by 4 pixel x4 line unit in a H.264 system.

  For example, a convex projection method (POCS) is known as a method of obtaining two or more constraints (parameters) alternately to obtain an appropriate result gradually. In the convex projection method, when a set satisfying two constraint conditions is a convex set, it is guaranteed to converge by repetition. The intra prediction apparatus according to the present invention uses a predetermined high-frequency region of an orthogonal transform coefficient of a predicted image in order to perform prediction with less influence of noise components and without losing continuity with decoded images including AC components. Two constraint conditions are used: correction to a value, and correction to an original value when a reference image adjacent to a predicted image is changed by calculation.

  For the prediction, not only a pixel adjacent to the prediction pixel but also a decoded pixel having a certain region is used, and orthogonal transformation such as DCT or FFT (Fast Fourier Transform) is used. Then, the projection is repeated to harmonize the two constraints. Since the prediction using a large number of adjacent reference pixels and emphasizing the low frequency component is performed, the influence of noise can be reduced.

  Specifically, the orthogonal transformation and the inverse orthogonal transformation are repeatedly performed on the block including the predicted image and the reference image while performing image correction to approximate the constraint condition. Ideally, as a result of repetition, the value of the pixel corresponding to the reference pixel becomes the value of the reference pixel, and the value of the pixel corresponding to the prediction pixel should be continuous with the value of the reference pixel in a direct current and alternating current manner. That's fine. However, there is no guarantee that it will converge in such a state, and even if it converges, the number of iterations will increase and the amount of computation will increase. Therefore, by gradually relaxing the constraint conditions, an effective result can be obtained with a reasonable amount of computation. Is preferred.

  Hereinafter, embodiments according to the present invention will be described in detail with reference to the drawings.

  FIG. 1 is a block diagram illustrating a configuration of an encoding-side intra prediction apparatus according to a first embodiment of the present invention. The intra prediction apparatus 1 includes a reference image analysis unit 10, a predicted image initial value setting unit 11, a processing block generation unit 12, an orthogonal transform unit 13, a coefficient correction unit 14, an inverse orthogonal transform unit 15, and a predicted image. An extraction unit 18, a control unit 19, and a control parameter transmission unit 20 are provided.

  When performing an iterative process described later, the intra prediction device 1 further includes a branching unit 16 and an image correcting unit 17. Even when the repetitive processing is not performed, a certain effect can be obtained. However, the discontinuity of the AC component can be further reduced by performing the repetitive processing. In the following description, a case where the iterative process is performed will be described.

  FIG. 2 is a diagram for explaining processing of the intra prediction apparatus 1 according to the first embodiment of the present invention. Specifically, FIG. 2A and FIG. 2C are diagrams illustrating a processing block P that is a unit of processing performed in the intra prediction apparatus 1. FIGS. 2B and 2D are diagrams showing the frequency region of the processing block P. The upper left region indicates a low frequency component, and the lower right region indicates a high frequency component. The image block X is an image block (hereinafter referred to as “prediction block”) that is a target of intra prediction. The image blocks D, B, and A are image blocks that have been encoded and then decoded, and are image blocks that are referred to in order to predict pixel values of the prediction block (hereinafter referred to as “reference blocks”). In the following description, the images in the reference blocks D, B, and A are referred to as reference images d, b, and a, respectively, and the image in the prediction block X is referred to as a predicted image x.

  The size of each block is set in advance. As shown in FIG. 2A, the reference block D is k pixels × m lines, the reference block B is 1 pixel × m lines, and the reference block A is k pixels × n lines. The prediction block X is 1 pixel × n lines, and the processing block P is (k + 1) pixels × (m + n) lines.

  The predicted image initial value setting unit 11 sets values obtained from all or part of the pixel values of the reference images d, b, and a as initial values of the predicted image x based on instructions from the control unit 19. For example, it is preferable that the average value of the pixel values of the reference images d, b, and a is the initial value because the DC component is retained. In addition, an average value of pixels adjacent to the prediction block X is used as an initial value, or the conventional MPEG4 AVC / H. The result of intra prediction by the H.264 method can be used as an initial value. Then, this initial value is output to the processing block generator 12.

  Based on the instruction from the control unit 19, the processing block generation unit 12 uses the reference blocks D, B, and A that are input and the initial value of the predicted image x input from the predicted image initial value setting unit 11, as shown in FIG. As shown in (a), a processing block P including reference blocks D, B, and A and a prediction block X in which initial values are set is generated and output to the orthogonal transform unit 13.

  The orthogonal transform unit 13 performs orthogonal transform on the image of the processing block P input from the processing block generation unit 12 and outputs the derived orthogonal transform coefficient to the coefficient correction unit 14.

  In the boundary region between the reference blocks D, B, A and the prediction block X, the coefficient correction unit 14 provides alternating continuity between the reference images d, b, a and the prediction image x (reducing discontinuity). Therefore, the high frequency component of the processing block P is suppressed based on an instruction from the control unit 19. Therefore, among the orthogonal transform coefficients input from the orthogonal transform unit 13, as shown in FIG. 2B, the absolute value of the orthogonal transform coefficient in the coefficient correction region H, which is a high frequency region, is set to a value smaller than the original value. Correct it. The number h of coefficients in the coefficient correction region H is 1 ≦ h <(k + 1) × (m + n). Then, the corrected orthogonal transform coefficient is output to the inverse orthogonal transform unit 15. A value to be corrected is preferably 0 in order to remove high frequency components. Further, the coefficient correction region H in which the orthogonal transformation is corrected may be a square shape or a rectangular shape as shown in FIG.

The coefficient correction unit 14 corrects the absolute values of the h _n orthogonal transform coefficients in order from the highest order orthogonal transform coefficient to a value smaller than the original value when the n-th process is performed by an iterative process described later. . Here, in order to gradually relax the constraint condition that the coefficient correction region H is corrected to a predetermined value, each time the process is repeated, the area of the coefficient correction region H after the repetition is made equal to or less than the area before the repetition, and h ₁ It is preferable that ≧ h ₂ ≧... ≧ h _n .

  The inverse orthogonal transform unit 15 performs an inverse orthogonal transform on the modified orthogonal transform coefficient input from the coefficient modification unit 14 to generate an image, and outputs the image to the branching unit 16.

  Based on an instruction from the control unit 19, the branching unit 16 determines whether or not to repeat a series of processes of image correction described later and the above-described orthogonal transform, orthogonal transform coefficient correction, and inverse orthogonal transform. For example, when the number of repetitions is less than or equal to a predetermined number, the image input from the inverse orthogonal transform unit 15 is output to the image correction unit 17 in order to execute the repetition process, and the number of repetitions exceeds the predetermined number. In this case, the iterative process is terminated and the image input from the inverse orthogonal transform unit 15 is output to the predicted image extraction unit 18.

  Further, as another determination method in the branching unit 16, for the regions in the reference blocks D, B, and A, a difference value between the current pixel value and the pixel value of the original decoded image before the iterative process is calculated. If the difference value is greater than or equal to the threshold value, the image input from the inverse orthogonal transform unit 15 is output to the image correction unit 17 in order to execute the repetition process. If the difference value is smaller than the threshold value, the repetition process ends. The image input from the inverse orthogonal transform unit 15 may be output to the predicted image extraction unit 18. The areas in the reference blocks D, B, and A for which the difference values are calculated are preferably image correction areas G that are corrected by the image correction unit 17 described below. Even in this determination method, it is preferable to set an upper limit number of repetitions in order to limit the number of repetitions.

  The image correction unit 17 is input from the branching unit 16 based on an instruction from the control unit 19 in order to provide alternating continuity between the original reference image d, b, a and the predicted image x in the boundary region. Correct the image that will be displayed. Discontinuity of pixel values in the boundary region between the reference block D, B, A and the prediction block X by correcting the absolute value of the higher-order orthogonal transform coefficient to a value smaller than the original value and performing inverse orthogonal transform However, the pixel values of the reference blocks D, B, and A are changed from the original values. Therefore, for example, as illustrated in FIG. 2C, the pixel values of the image correction region G adjacent to the prediction block X in the reference blocks D, B, and A are changed to the original values of the reference blocks D, B, and A. Correct to pixel value. The number of pixels g in the image correction area G is 1 ≦ g ≦ k × m + 1 × m + k × n. Then, the processing block P after correcting the pixel value is output to the orthogonal transform unit 13.

Image correcting unit 17, when performing n-th process in the iteration, see block D, B, of the A, to correct the image correction region G made of g _n pixels to the original pixel values. Here, in order to gradually relax the constraint condition that the reference image d, b, a is corrected to the original value when the reference image d, b, a is changed by calculation, the area of the image correction region G after the repetition is repeated each time the processing is repeated. repeating the following previous area, it is preferable to the _{g 1 ≧ g 2 ≧ ... ≧} g n.

  The predicted image extraction unit 18 extracts an image of the region of the prediction block X from the image input from the branching unit 16 and sets it as a predicted image for intra prediction.

  The reference image analysis unit 10 analyzes the frequencies of the input reference images d, b, and a and outputs a determination signal to the control unit 19.

  FIG. 3 is a block diagram illustrating a configuration relating to control based on reference image analysis of the intra prediction apparatus 1, in which a configuration of the reference image analysis unit 10 and a partial configuration of the control unit 19 are illustrated. The reference image analysis unit 10 includes an orthogonal transform unit 101 (101-A, 101-B, 101-D), a high frequency extraction unit 102 (102-A, 102-B, 102-D), a power calculation unit 103, And a power determination unit 104.

  The orthogonal transform unit 101-A performs orthogonal transform of the reference image a, and outputs the derived orthogonal transform coefficient to the high frequency extraction unit 102-A. Similarly, the orthogonal transform unit 101-B performs orthogonal transform of the reference image b, outputs the derived orthogonal transform coefficient to the high frequency extraction unit 102-B, and the orthogonal transform unit 101-D performs orthogonal transform of the reference image d. And the derived orthogonal transform coefficient is output to the high frequency extraction unit 102-D.

  The high frequency extraction unit 102 -A extracts a high frequency component from the orthogonal transformation coefficient input from the orthogonal transformation unit 101 -A and outputs the high frequency component to the power calculation unit 103. Similarly, the high frequency extraction unit 102-B extracts a high frequency component from the orthogonal transformation coefficient input from the orthogonal transformation unit 101-B and outputs the high frequency component to the power calculation unit 103. The high frequency extraction unit 102-D includes the orthogonal transformation unit. A high frequency component is extracted from the orthogonal transform coefficient input from 101 -D and output to the power calculation unit 103. When the block sizes of the reference blocks A, B, D and the prediction block X do not match, normalization processing is performed for the calculation. Here, the high frequency component is not limited to the value of the orthogonal transformation coefficient of a component higher than a certain frequency, and may be any index that represents the high frequency component of the image. For example, a value obtained by weighted addition of orthogonal transform coefficients of various frequency components by a preset method may be used.

  The power calculation unit 103 adds the high-frequency components input from the high-frequency extraction units 102-A, 102-B, and 102-D, adds the values other than 0, or adds the weights, etc., or the entire reference images d, b, and a. Is derived and output to the power determination unit 104.

  The power determination unit 104 determines the magnitude of the high frequency components of the entire reference images a, b, and d input from the power calculation unit 103 by comparing with a threshold value, and outputs a determination signal to the control unit 19. A plurality of threshold values may be provided. The determination signal is, for example, a 1-bit signal that indicates the magnitude of power in two levels, that is, when the threshold value is one, and when the threshold value is two, the magnitude of the power is large. It is a 2-bit signal shown in three stages of medium and small.

  Instead of the orthogonal transform unit 101 and the high-frequency extraction unit 102, high-frequency components can be extracted using a high-pass filter or a filter that combines a high-pass filter, a band-pass filter, and a low-pass filter.

  The control unit 19 sets a parameter that defines an initial value of the predicted image x (whether the initial value is an average value or an intra prediction result according to the MPEG4 AVC / H.264 method) to the predicted image initial value setting unit 11. And outputs the parameters (k, l, m, n) that define the regions of the reference blocks D, B, A and the prediction block X to the processing block generation unit 12, and the parameters (of H Size) and a parameter that defines the correction value of the orthogonal transform coefficient are output to the coefficient correction unit 14, and a parameter that specifies the repetition process (the number of repetitions and / or the threshold value of the reference block difference value) is output to the branching unit 16. A parameter (G size) that defines the image correction region G is output to the image correction unit 17. Further, the set control parameter is output to the control parameter transmission unit 20.

  Referring to FIG. 3, the control unit 19 includes a coefficient correction area control unit 191 and an image correction area control unit 192. The coefficient correction area control unit 191 determines a parameter (H size) that defines the coefficient correction area H based on the determination signal input from the power determination unit 104 and outputs the parameter to the coefficient correction unit 14. When the high frequency power of the reference blocks D, B, A is small, the high frequency power of the processing block P is also small, and the high frequency component is mainly caused by discontinuity between the reference images d, b, a and the predicted image x. It is believed that there is. Therefore, the coefficient correction area control unit 191 determines parameters so that the coefficient correction area H is increased when the high frequency power of the reference blocks D, B, and A is small compared to when the high frequency power is large.

  The image correction area control unit 192 determines a parameter (G size) that defines the image correction area G based on the determination signal input from the power determination unit 104 and outputs the parameter to the image correction unit 17. When the high frequency power of the reference blocks D, B, and A is small, it is considered that there is little change in the pattern in the processing block P. Therefore, the image correction area control unit 192 determines the parameter so that the image correction area G is larger when the high frequency power of the reference blocks D, B, and A is smaller than when the high frequency power is large.

  For example, when the determination signal is a signal indicating the magnitude of power in n stages, the coefficient correction area control unit 191 prepares n types of parameters that define the coefficient correction area H in advance. A parameter that makes the region H large is selected. Similarly, the image correction area control unit 192 prepares n types of parameters that define the image correction area G in advance, and selects a parameter that increases the size of the image correction area G as the power decreases.

  Here, the parameter is not limited to the size of the coefficient correction area H or the image correction area G, but may be an increase / decrease value with respect to a standard size set for each repetition count. Alternatively, the parameter may be determined such that the larger the power, the larger the amount to be reduced for each repetition of the coefficient correction area H and the image correction area G.

  In addition, although the reference image analysis unit 10 has described the example in which the analysis is performed by distinguishing the frequency into two stages of the high frequency and the others, the analysis may be performed by distinguishing the frequency into a plurality of stages. Further, when the high-frequency power is equal to or higher than a predetermined threshold value, the coefficient correction area H and / or the image correction area G may not be set.

  The control parameter transmission unit 20 transmits the control parameters input from the control unit 19 to the outside. Further, as described above, the predicted image initial value setting unit 11 performs the conventional MPEG4 AVC / H. When the result of intra prediction by the H.264 method is used as the initial value of the predicted image x, the conventional intra prediction mode needs to be transmitted as parameter information. In addition, when the parameter which prescribes | regulates the initial value of the prediction image x, the parameter which prescribes | regulates the area | region of reference block D, B, A, and the prediction block X, and the parameter which prescribes | regulates an iterative process are made into a fixed value, these parameters are There is no need for transmission, and the control parameter transmission unit 20 is not necessary.

  FIG. 4 is a block diagram illustrating a configuration of the decoding-side intra prediction device 1 ′ according to the first embodiment of the present invention. The intra prediction apparatus 1 ′ on the decoding side is different from the intra prediction apparatus 1 on the encoding side in that it does not include the control parameter transmission unit 20 but includes the control parameter reception unit 21.

  The control parameter receiving unit 21 receives control parameters from the control parameter transmission unit 20 of the intra prediction apparatus 1 on the encoding side, and outputs the received control parameters to the control unit 19. The control unit 19 determines a parameter determined based on the control parameter input from the control parameter receiving unit 21 and the determination signal input from the reference image analysis unit 10, the predicted image initial value setting unit 11, the processing block generation unit 12, the coefficient The correction unit 14, the branch unit 16, and the image correction unit 17 are set. In addition, when the parameter that defines the initial value of the predicted image x, the parameter that defines the areas of the reference blocks D, B, and A and the predicted block X, and the parameter that defines the repetition process are fixed values, the control parameter receiving unit 21 becomes unnecessary. The processing in each of the other processing units is the same as that of the intra device 1 on the encoding side, and thus description thereof is omitted.

  Even in the intra prediction apparatus 1 ′ on the decoding side, in the same manner as the intra prediction apparatus 1 on the encoding side, the reference image analysis unit 10 analyzes the reference image, so that the amount of transmission information from the encoding side is not increased. Parameter settings suitable for the image can be performed.

[Operation of intra prediction device]
Next, the operation of the intra prediction apparatus will be described with reference to FIG. FIG. 5 is a flowchart showing the operation of the intra prediction apparatus 1 (1 ′) according to the first embodiment of the present invention.

  In step S101, the reference image analysis unit 10 analyzes the frequencies of the input reference images d, b, and a.

  In step S102, the predicted image initial value setting unit 11 obtains an initial value of the predicted image x and sets it to the predicted block X.

  In step S103, the processing block generation unit 12 generates a processing block P including the reference blocks D, B, and A and the prediction block X in which the initial value is set.

  In step S104, the orthogonal transform unit 13 performs orthogonal transform on the image of the processing block P generated in step S103 to derive orthogonal transform coefficients.

  In step S105, the coefficient correction unit 14 corrects h orthogonal transform coefficients to 0 in order from the highest order among the orthogonal transform coefficients derived in step S104.

  In step S106, the inverse orthogonal transform unit 15 performs inverse orthogonal transform on the orthogonal transform coefficient corrected in step S105 to generate an image.

  In step S107, the branching unit 16 determines whether or not the processing from step S104 to step S106 has been repeated a predetermined number of times. If it is determined that the repetition has not reached the predetermined number of times, the process proceeds to step S108, and if it is determined that the repetition has reached the predetermined number of times, the process proceeds to step S109. Note that the determination method may be a method of determining a threshold value as described above.

  In step S108, the image correction unit 17 corrects g pixel values adjacent to the prediction block X among the reference blocks D, B, and A to the original reference pixel values for the image generated in step S106. Then, the process returns to step S103.

  In step S109, the predicted image extracting unit 18 extracts a region of the predicted block X from the image generated in step S106 to generate a predicted image.

[Encoder and decoder]
FIG. 6 is a block diagram illustrating a configuration of the encoder 2 including the intra prediction device 1 according to the first embodiment of the present invention. The encoder 2 is a conventional MPEG4 AVC / H. Instead of the intra prediction unit in the H.264 encoder (see FIG. 9), an intra prediction device 1 according to the present invention is provided. The intra prediction apparatus 1 uses a reference image analysis unit 10, a predicted image initial value setting unit 11, a processing block generation unit 12, and an image obtained by decoding an already-encoded block via an inverse quantization unit 35 and an inverse orthogonal transform unit 36. The prediction image x that has been input to the correction unit 17 and generated is output to the subtraction unit 31 and the addition unit 41 via the changeover switch 37. The control parameter transmission unit 20 outputs the set control parameter to the variable length coding unit 34.

  FIG. 7 is a block diagram illustrating a configuration of the decoder 3 including the decoding-side intra prediction device 1 ′ according to the first embodiment of the present invention. The decoder 3 includes a variable length decoding unit 51, an inverse quantization unit 52, an inverse orthogonal transform unit 53, an addition unit 54, an intra prediction device 1 ′, a frame memory 55, a motion compensation prediction unit 56, And a changeover switch 57. The decoder 3 is a conventional MPEG4 AVC / H. Instead of the intra prediction unit of the H.264 decoder, an intra prediction device 1 ′ according to the present invention is provided.

  The variable length decoding unit 51 receives a bit stream encoded by inter-frame prediction, performs variable length decoding processing, outputs the bit stream to the inverse quantization unit 52, and decodes motion vector information to perform motion compensation prediction unit To 56. Further, when control parameter information is transmitted from the encoder side, the control parameter information is decoded and output to the control parameter receiving unit 21.

  The inverse quantization unit 52 performs an inverse quantization process on the quantized orthogonal transform coefficient input from the variable length decoding unit 51 to obtain an orthogonal transform coefficient of the motion compensated difference image, and performs an inverse orthogonal transform To the unit 53.

  The inverse orthogonal transform unit 53 performs inverse orthogonal transform (for example, IDCT) on the orthogonal transform coefficient of the difference image input from the inverse quantization unit 52, and outputs the obtained difference image to the addition unit 54.

  The adding unit 54 restores the image by adding the difference image obtained from the inverse orthogonal transform unit 53 and the predicted image input from the motion compensated prediction unit 56 or the predicted image x input from the intra prediction device 1 ′. And output to the intra prediction device 1, the frame memory 55, and the outside.

  The motion compensation prediction unit 56 generates a prediction image using the reference image obtained from the frame memory 55 and the motion vector obtained from the variable length decoding unit 51, and outputs the prediction image to the addition unit 54 via the changeover switch 57.

  The intra prediction apparatus 1 ′ receives an image decoded through the inverse quantization unit 52 and the inverse orthogonal transform unit 53, and as described above, the predicted image initial value setting unit 11, the processing block generation unit 12, and the orthogonal transform unit 13 are input. The coefficient correction unit 14, the inverse orthogonal transform unit 15, the branching unit 16, the image correction unit 17, the predicted image extraction unit 18, and the control unit 19 generate a predicted image x, and the predicted image x is switched via the changeover switch 57. The result is output to the adder 54.

  Thus, according to the intra prediction apparatus 1 of Example 1, the power of the reference images d, b, and a is estimated, and the size of the coefficient correction region H and the size of the image correction region G are set based on the estimation result. By doing so, it becomes possible to generate the predicted image x according to the frequency characteristics of the reference images d, b, and a.

  Next, the intra prediction apparatus 1 of Example 2 by this invention is demonstrated. The same components as those in the first embodiment are denoted by the same reference numerals, and description thereof is omitted. Compared with the intra prediction device 1 of the first embodiment, the intra prediction device 1 of the second embodiment only performs processing based on the signal input from the reference image analysis unit 10 and the reference image analysis unit 10 of the control unit 19. Is different.

  FIG. 8 is a block diagram illustrating a configuration related to control based on reference image analysis of the intra prediction apparatus 1 according to the second embodiment, in which a configuration of the reference image analysis unit 10 and a partial configuration of the control unit 19 are illustrated. Yes. The reference image analysis unit 10 includes an orthogonal transform unit 101 (101-A, 101-B, 101-D), a high frequency extraction unit 102 (102-A, 102-B, 102-D), and a horizontal high frequency estimation unit 105. And a vertical high-frequency estimation unit 106.

  The orthogonal transform unit 101-A outputs the orthogonal transform coefficient of the reference image a to the high frequency extraction unit 102-A, and the orthogonal transform unit 101-B outputs the orthogonal transform coefficient of the reference image b to the high frequency extraction unit 102-B. The orthogonal transform unit 101-D outputs the orthogonal transform coefficient of the reference image d to the high frequency extraction unit 102-D.

The high frequency extraction unit 102-A extracts a high frequency component from the orthogonal transformation coefficient input from the orthogonal transformation unit 101-A, and outputs a horizontal high frequency component (horizontal high frequency component) _PAH to the horizontal high frequency estimation unit 105. and outputs the vertical high-frequency component (vertical high-frequency component) P _AV vertically high-frequency estimating unit 106. The high frequency extraction unit 102-B extracts a high frequency component from the orthogonal transformation coefficient input from the orthogonal transformation unit 101-B, outputs the horizontal high frequency component P _BH to the horizontal high frequency estimation unit 105, and vertically converts the vertical high frequency component P _BV . It outputs to the high frequency estimation part 106. The high frequency extraction unit 102-D extracts a high frequency component from the orthogonal transform coefficient input from the orthogonal transform unit 101-D, outputs the horizontal high frequency component _PDH to the horizontal high frequency estimation unit 105, and vertically converts the vertical high frequency component _PDV . It outputs to the high frequency estimation part 106. Here, the high frequency component is not limited to the value of the orthogonal transformation coefficient of a component higher than a certain frequency, and may be any index that represents the high frequency component of the image. For example, a value obtained by weighted addition of orthogonal transform coefficients of various frequency components by a preset method may be used.

The horizontal high frequency estimator 105 calculates the horizontal high frequency estimation amount P _{XH of the} prediction block X based on the horizontal high frequency components P _AH, P _BH , and P _DH input from the high frequency extractors 102 -A, 102 -B, and 102 -D. Is calculated and output to the control unit 19. The reference block A and the prediction block X are adjacent in the horizontal direction, and it is appropriate that the horizontal property of the prediction block X is estimated by the reference block A. Further, the relationship between the reference block A and the prediction block X can be estimated from the relationship between the reference block D and the reference block B. For example, even when both the reference block D and the reference block A have a large horizontal high-frequency component, if the reference block B has a smaller high-frequency component than the reference block D, the prediction block X also has a smaller high-frequency component than the reference block A. Is estimated. Therefore, the horizontal high frequency estimation amount P _XH is calculated by the following equation (1), for example.
P _XH = P _AH + α × (P _BH −P _DH ) (1)
Here, α is a coefficient that can be arbitrarily set. For example, α = P _AH / P _DH or α = 1.

Similarly, the vertical high frequency estimation unit 106 estimates the vertical high frequency of the prediction block X based on the vertical high frequency components P _AV, P _BV , and P _DV input from the high frequency extraction units 102-A, 102-B, and 102-D. The quantity P _XV is calculated by the following equation (2), for example, and output to the control unit 19.
P _XV = P _BV + β × (P _AV −P _DV ) (2)
Here, β is a coefficient that can be arbitrarily set. For example, β = P _BV / P _DV or β = 1.

  Instead of the orthogonal transform unit 101 and the high-frequency extraction unit 102, high-frequency components can be extracted using a high-pass filter or a filter that combines a high-pass filter, a band-pass filter, and a low-pass filter.

The coefficient correction area control unit 191 defines the coefficient correction area H based on the horizontal high-frequency estimation amount P _XH input from the horizontal high-frequency estimation unit 105 and the vertical high-frequency estimation amount P _XV input from the vertical high-frequency estimation unit 106. A parameter (size of H) is determined and output to the coefficient correction unit 14. When the horizontal high frequency estimation amount _PXH is small, it can be estimated that the horizontal correlation of the prediction block X is high. Therefore, the coefficient correction area control unit 191 prepares a plurality of types of parameters that define the horizontal width of the coefficient correction area H in advance, and the smaller the horizontal high-frequency estimation amount _PXH is, the smaller the horizontal direction of the coefficient correction area H is. Select a parameter to increase the width. Further, when the vertical high-frequency estimation amount P _XV is small, it can be estimated that the vertical correlation of the prediction block X is high. Therefore, the coefficient correction area control unit 191 prepares a plurality of types of parameters that define the length of the coefficient correction area H in the vertical direction in advance, and the vertical direction of the coefficient correction area H decreases as the vertical high-frequency estimation amount P _XV decreases. Select a parameter to increase the length of.

The image correction area control unit 192 defines the image correction area G based on the horizontal high-frequency estimation amount P _XH input from the horizontal high-frequency estimation unit 105 and the vertical high-frequency estimation amount P _XV input from the vertical high-frequency estimation unit 106. The parameter (G size) is determined and output to the image correction unit 17. When the horizontal high frequency estimation amount _PXH is small, it can be estimated that the horizontal correlation between the adjacent reference blocks D, B, and A is also high. Therefore, the image correction area control unit 192 prepares a plurality of types of parameters that define the horizontal width of the image correction area G in advance. The smaller the horizontal high-frequency estimation amount _PXH is, the more the horizontal direction of the image correction area G is. Select a parameter to increase the width. Further, when the vertical high-frequency estimation amount P _XV is small, it can be estimated that the vertical correlation between the adjacent reference blocks D, B, and A is also high. Therefore, the image correction area control unit 192 prepares in advance a plurality of types of parameters that define the length of the image correction area G in the vertical direction, and the vertical direction of the image correction area G decreases as the vertical high-frequency estimation amount P _{XV decreases.} Select a parameter to increase the length of.

Note that the parameter may not be the size itself of the coefficient correction area H or the image correction area G, but may be an increase / decrease value with respect to the standard size set for each repetition count. The parameter is determined such that the larger the horizontal high-frequency component _PXH is, the larger the amount to be reduced for each repetition of the horizontal width of the coefficient correction area H and the horizontal width of the image correction area G is determined. The parameter may be determined such that the larger the value P _XV is, the larger the amount of reduction is repeated for each repetition of the vertical length of the coefficient correction area H and the vertical length of the image correction area G. Further, when the horizontal high-frequency component P _XH and the vertical high-frequency component P _XV are equal to or greater than a predetermined threshold value, the coefficient correction region H may not be set, that is, the orthogonal transform coefficient may not be corrected.

  As described above, according to the intra prediction apparatus 1 of the second embodiment, the horizontal power and the vertical power of the reference images d, b, and a are estimated, and the horizontal direction of the coefficient correction region H is estimated based on the estimation result. By setting the width and the length in the vertical direction, and the horizontal width and the length in the vertical direction of the image correction region G, a more optimal predicted image x is generated according to the frequency characteristics of the reference images d, b, and a. Will be able to.

  Here, in order to function as the intra prediction apparatus 1 on the encoding side, a computer can be preferably used, and by causing the computer to execute a predetermined program by the CPU, the reference image analysis unit 10 and the predicted image initial stage A value setting unit 11, a processing block generation unit 12, an orthogonal transformation unit 13, a coefficient modification unit 14, an inverse orthogonal transformation unit 15, a branching unit 16, an image modification unit 17, a predicted image extraction unit 18, The functions of the control unit 19 and the control parameter transmission unit 20 can be realized.

  Similarly, in order to function as the intra prediction apparatus 1 ′ on the decoding side, a computer can be suitably used. By causing the computer to execute a predetermined program, the predicted image initial value setting unit 11 and the processing are performed. Block generation unit 12, orthogonal transformation unit 13, coefficient modification unit 14, inverse orthogonal transformation unit 15, branching unit 16, image modification unit 17, predicted image extraction unit 18, control unit 19, and control parameters The function of the receiving unit 21 can be realized.

  Similarly, in order to function as the encoder 2, a computer can be suitably used. By causing the computer to execute a predetermined program by the CPU, the code-side intra prediction device 1, the subtracting unit 31, An orthogonal transform unit 32, a quantization unit 33, a variable length coding unit 34, an inverse quantization unit 35, an inverse orthogonal transform unit 36, a changeover switch 37, a frame memory 39, a motion compensation prediction unit 40, The function of the adding unit 41 can be realized.

  Similarly, in order to function as the decoder 3, a computer can be suitably used. By causing the computer to execute a predetermined program, the intra prediction device 1 on the decoding side, the variable length decoding unit 51, The functions of the inverse quantization unit 52, the inverse orthogonal transform unit 53, the addition unit 54, the frame memory 55, the motion compensation prediction unit 56, and the changeover switch 57 can be realized.

  Although the above embodiments have been described as representative examples, it will be apparent to those skilled in the art that many changes and substitutions can be made within the spirit and scope of the invention. Therefore, the present invention should not be construed as being limited by the above-described embodiments, and various modifications and changes can be made without departing from the scope of the claims. For example, when performing intra prediction, prediction based on the above-described method and conventional MPEG-4 AVC / H. The prediction by the H.264 system may be switched, or a prediction image by both methods may be generated to determine and output a prediction image with good coding efficiency.

  As described above, according to the present invention, it is possible to perform intra prediction suitable for various images, which is useful for any application for encoding or decoding an image.

1, 1 ′ Intra Prediction Device 10 Reference Image Analysis Unit 11 Predicted Image Initial Value Setting Unit 12 Processing Block Generation Unit 13 Orthogonal Transform Unit 14 Coefficient Correction Unit 15 Inverse Orthogonal Transform Unit 16 Branching Unit 17 Image Correction Unit 18 Predictive Image Extraction Unit 19 Control unit 20 Control parameter transmission unit 21 Control parameter reception unit 101-A, 101-B, 101-D Orthogonal transformation unit 102-A, 102-B, 102-D High frequency extraction unit 103 Power calculation unit 104 Power determination unit 105 Horizontal High-frequency estimation unit 106 Vertical high-frequency estimation unit 191 Coefficient correction region control unit 192 Image correction region control unit 2 Encoder 3 Decoder

Claims (9)

An intra prediction device that generates a prediction image of a prediction block from an image of a decoded reference block adjacent to the prediction block,
A predicted image initial value setting unit that sets an initial value of the predicted image in a predicted block;
A processing block generation unit that generates a processing block including a reference block and a prediction block in which the initial value is set;
An orthogonal transform unit that performs orthogonal transform on the image of the processing block to generate an orthogonal transform coefficient;
A coefficient correction unit that corrects the absolute value of the orthogonal transform coefficient generated by the orthogonal transform unit to a value smaller than the original value in a high-frequency region;
An inverse orthogonal transform unit that generates an image by performing inverse orthogonal transform on the modified orthogonal transform coefficient;
An image correcting unit that corrects an image area in the reference block of the image generated by the inverse orthogonal transform unit to a pixel value of an original decoded image, and supplies the corrected image to the orthogonal transform unit; ,
A predicted image extraction unit that extracts a predicted block image from the image generated by the inverse orthogonal transform unit and generates a predicted image;
It is determined whether or not a predetermined condition for ending the repetition of a series of processes by the image correction unit, the orthogonal transform unit, the coefficient correction unit, and the inverse orthogonal transform unit is satisfied, and the predetermined condition is not satisfied In the case where it is determined that the image generated by the inverse orthogonal transform unit is supplied to the image correction unit in order to repeatedly execute the series of processing, and it is determined that the predetermined condition is satisfied A branching unit that finishes repeating the series of processes and supplies the image generated by the inverse orthogonal transform unit to the predicted image extraction unit;
A reference image analysis unit for analyzing the frequency of the image of the reference block;
A control unit that determines at least one of a high-frequency region in which the coefficient correction unit corrects an orthogonal transform coefficient, or an image region in which the image correction unit corrects a pixel value, based on an analysis result obtained by the reference image analysis unit. When,
An intra prediction apparatus comprising: The reference image analysis unit derives the power of the high frequency component of the reference block,
The control unit determines that at least one of a high-frequency region in which the coefficient correction unit corrects an orthogonal transform coefficient or an image region in which the image correction unit corrects a pixel value increases as the derived power decreases. The intra prediction apparatus according to claim 1, wherein: The reference image analysis unit derives the power of the high-frequency component in the horizontal direction and the power of the high-frequency component in the vertical direction of the reference block,
As the power of the derived horizontal high-frequency component is smaller, the control unit has a horizontal width of a high-frequency region in which the coefficient correction unit corrects an orthogonal transform coefficient, or an image in which the image correction unit corrects a pixel value. It is determined that at least one of the horizontal widths of the region is long and the power of the derived vertical high-frequency component is small, the coefficient correction unit corrects the orthogonal transform coefficient in the vertical direction of the high-frequency region. The intra prediction apparatus according to claim 1, wherein at least one of a length or a length in a vertical direction of an image region whose pixel value is corrected by the image correction unit is determined to be longer.   The coefficient correction unit corrects the absolute value of the orthogonal transform coefficient generated by the orthogonal transform unit to a value smaller than the original value in a high-frequency region including the highest-order orthogonal transform coefficient, and each time it is repeatedly executed. The intra prediction apparatus according to any one of claims 1 to 3, wherein an area of the high-frequency region for correcting the orthogonal transform coefficient is equal to or less than an area before the repetition.   The image correction unit corrects an image area including an image adjacent to the prediction block in the reference block to a pixel value of an original decoded image, and corrects the pixel value each time the execution is repeated. The intra prediction apparatus according to any one of claims 1 to 4, wherein an area of the image region is equal to or less than an area before the repetition.   The predetermined condition for terminating the repetition of the series of processes is that the number of the series of processes is equal to or greater than the predetermined number, or an image of the reference block after the series of processes and an image of the reference block before the series of processes are performed. The intra prediction apparatus according to claim 1, wherein the difference between the pixel values is equal to or less than a predetermined threshold.   The encoder provided with the intra prediction apparatus as described in any one of Claim 1 to 6.   A decoder comprising the intra prediction device according to any one of claims 1 to 6. A computer functioning as an intra prediction device that generates a predicted image of a prediction block from an image of a decoded reference block adjacent to the prediction block,
(A) analyzing the frequency of the image of the reference block;
(B) setting an initial value of the predicted image in the predicted block;
(C) generating a processing block including the reference block and a prediction block in which the initial value is set;
(D) performing orthogonal transform on the image of the processing block to generate orthogonal transform coefficients;
(E) correcting the absolute value of the orthogonal transform coefficient generated by the step (d) to a value smaller than the original value in a high-frequency region;
(F) performing an inverse orthogonal transform on the orthogonal transform coefficient modified in the step (e) to generate an image;
(G) determining whether or not a predetermined condition for ending the repetition of a series of processes from step (d) to step (f) is satisfied;
(H) If it is determined in step (g) that the predetermined condition is not satisfied, among the images generated in step (f), an image area in the reference block is already decoded. Correcting the pixel value of the image of step (d) and repeating step (g) from step (d);
(I) When it is determined in step (g) that the predetermined condition is satisfied, the series of processing is terminated and the predicted block image is generated from the image generated in step (f). Extracting a prediction image to generate a prediction image;
(J) Based on the analysis result obtained in step (a), at least one of a high-frequency region for correcting the orthogonal transform coefficient in step (e) and an image region for correcting the pixel value in step (h) A step of determining
A program for running

Images