JP2008072608A - Apparatus and method for encoding image - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To sharply project a designated object in an image while taking an entire processing load into account. <P>SOLUTION: An image encoding apparatus includes: an object region management means for managing an object region indicating a position of a designated object in a designated image; a peripheral region determining means for determining a peripheral region including the object region of the designated object in the image; a search range control means for controlling to widen a search area of a motion search for the peripheral region determined by the peripheral region determining means and to narrow the search area of the motion search for regions other than the peripheral region; a motion vector detecting means for detecting a motion vector within the search area controlled by the search range control means; an object region updating means for updating the object region of the designated object in the image based on the motion vector detected by the motion vector detecting means and the peripheral region; and an encoding control means for encoding the object region of the designated object in the image managed by the object region management means while setting a small quantization step size thereto. <P>COPYRIGHT: (C)2008,JPO&INPIT

Description

  The present invention relates to an image encoding device and an image encoding method, for example, an image encoding device that performs encoding processing on moving image data and outputs encoded image data in order to compress a moving image, and an image It can be applied to an encoding method.

  In order to compress moving images, a standardization technique standardized by MPEG is generally used as an encoding method for moving image data. The MPEG standard image encoding method divides each frame of a moving image into small blocks and performs encoding processing in units of blocks called macroblocks. The encoding processing is roughly divided into intra encoding and inter encoding. It is classified into two types.

  Intra coding is a compression process that uses the correlation between adjacent pixels in each frame of a moving image. In the intra coding method, one frame of an input image is divided into blocks, subjected to discrete cosine transform (DCT) for each block, quantized, and variable length coding is performed on the quantized DCT coefficients.

  Inter coding is a process of coding a difference signal between frames of a moving image. The inter coding method obtains an optimal prediction macroblock by performing motion compensation on the basis of a motion vector obtained from an encoding target image to be encoded and a reference image, and the prediction macroblock and a macro to be encoded. Find the difference signal from the block. The differential signal is quantized after being subjected to DCT, and the quantized DCT coefficient is variable-length encoded together with the motion vector and the quantization step size.

  Here, searching for the most similar block between the respective blocks of the encoding target image and the reference image is referred to as motion search, and the shift in the time direction of the position of the block obtained by the motion search is referred to as a motion vector. .

  Conventionally, as shown in Patent Documents 1 and 2, for example, as shown in Patent Documents 1 and 2, a motion search method in an image encoding device provides a fixed search range at each image position, or, for example, determines the size of a surrounding motion vector. The search range is determined accordingly, and the motion is detected by searching for the most similar block in the search range thus determined.

  In addition, when the DCT coefficient is quantized in the image coding apparatus, the quantization step size indicating the quantization width is constant for all blocks divided for each frame of the moving image, It is determined according to the code amount allocated to the image and each coefficient value.

JP 2004-140794 A JP 2003-274410 A

  By the way, when an image is encoded, it may be strongly desired that only a specific designated area is clearly displayed. For example, in the process of encoding a moving image such as a monitoring video, it is possible to improve the image quality of the monitoring target in the video and facilitate the identification of the monitoring target.

  However, in the above-described conventional encoding device, when the search range is widened, the calculation amount and the code amount increase accordingly, so that the hardware processing load differs. For this reason, the search range is limited according to the performance of the hardware. For example, when a wide search range is required, the motion search may not be successful, and the target to be noticed may be unclear.

  Also, in the DCT coefficient quantization process, the DCT coefficient is divided by the quantization step size to round down the remainder in order to reduce the code amount, but at the time of decoding, the quantization step size is added to the quantization value. Since it is restored by multiplying, an accurate value cannot be restored and distortion occurs. This distortion becomes smaller as the quantization step size becomes smaller.

  However, since the quantization step size is defined by the code amount that can be used as described above, the target to be noticed does not always appear clearly.

  Therefore, there is a need for an image encoding device and an image encoding method that can perform motion search of a target specified in an image more successfully than other portions, and can allocate a larger amount of code than other portions. Yes.

  In order to solve such a problem, an image encoding device according to a first aspect of the present invention is an image encoding device that encodes a moving image. (1) A target region indicating a position of a specified target in a specified image is displayed. A target area management means for managing; (2) a peripheral area determining means for determining a peripheral area including a target area to be specified in an image; and (3) a peripheral area determined by the peripheral area determining means. Search range control means for controlling the search range for motion search so that the search range for motion search is expanded and the search range for motion search is narrowed for areas other than the surrounding area; and (4) within the search range controlled by the search range control means. Motion vector detection means for detecting a motion vector in (5), and (5) target area update means for updating the target area to be designated in the image based on the motion vector detected by the motion vector detection means and the peripheral area; (6) to the target area of the specified object in the image managed by the target area management means, characterized in that it comprises a coding control means for coding is set smaller quantization step size.

  The image encoding method according to the second aspect of the present invention is the image encoding method for encoding a moving image, wherein (1) the target area management means manages a target area indicating the position of the specified target in the specified image. A target region management step, (2) a peripheral region determination step in which the peripheral region determination unit determines a peripheral region including the target region to be specified in the image, and (3) a search range control unit includes a peripheral region determination unit. (4) a search range control step for performing control so that the search range for motion search is expanded for the peripheral region determined by the above, and the search range for motion search is narrowed for regions other than the peripheral region; A motion vector detection step in which the motion vector detection means detects a motion vector within the search range controlled by the search range control means; and (5) a motion vector detected by the motion vector detection means by the target region update means. A target area update step of updating the target area to be specified in the image based on the peripheral area, and (6) the encoding control means for the target area to be specified in the image managed by the target area management means And an encoding control step for encoding with a small quantization step size.

  According to the present invention, it is possible to allocate a large amount of calculation and code necessary for encoding a specified area in an image, so that the specified area can be coded more clearly while suppressing the calculation amount and code amount as a whole. And motion information of a designated area can be provided.

(A) First Embodiment Hereinafter, a first embodiment of an image encoding device of the present invention will be described with reference to the drawings.

  In the first embodiment, a case where the image coding apparatus of the present invention is applied to an image coding apparatus that performs coding processing on image data to compress an input image will be described.

(A-1) Configuration of the First Embodiment FIG. 1 is a block diagram showing the main configuration of the image encoding device 100 of the first embodiment. As illustrated in FIG. 1, an image encoding device 100 according to the first embodiment includes an object management unit 1, a search range control unit 2, an encoding control unit 3, an mv detection unit 4, an inter prediction unit, and an intra prediction unit 6. , Block division unit 7, orthogonal transform unit 8, quantization unit 9, entropy encoding unit 10, inverse quantization unit 11, orthogonal inverse transform unit 12, mode selection unit 13, frame buffer 14, buffer management unit 15, frame buffer 16.

  In the first embodiment, the functions of the object management unit 1, the search range control unit 2, the encoding control unit 3, the mv detection unit 4, and the inter prediction unit 5 are different from those of the conventional image encoding device. The configuration other than these corresponds to the configuration of the conventional image encoding device.

  The block dividing unit 7 takes in an input image and divides the input image into blocks (in the first embodiment, a block of a rectangular pixel set of 16 pixels vertically and horizontally). The block dividing unit 7 gives the input image that has been made into a block to the orthogonal transform unit 8 and the mv detection unit 4.

  The orthogonal transform unit 8 calculates a difference between a predicted image (a block of a reference image that is most similar to an input image block in the case of inter prediction, and a pixel value calculated using peripheral pixels on the left and upper sides of the block in the case of intra prediction). Are orthogonally transformed. The orthogonal transform unit gives the transform coefficient obtained by frequency transform to the quantization unit 9.

  The quantization unit 9 quantizes the transform coefficient frequency-transformed by the orthogonal transform unit 8 and gives the quantized quantized value to the inverse quantization unit 11 and the entropy coding unit 10.

  The entropy encoding unit 10 performs entropy encoding on the transform coefficient, motion vector, and the like quantized by the quantization unit 9.

  The inverse quantization unit 11 performs inverse quantization on the image quantized by the quantization unit 9 in order to create a local decoded image to be used as a reference image by the inter prediction unit 5. The inverse quantization unit 11 gives the inversely quantized signal to the orthogonal inverse transform unit 12.

  The orthogonal inverse transform unit 12 performs orthogonal inverse transform on the signal inversely quantized by the inverse quantization unit 11 in order to obtain a local decoded image. Further, the orthogonal inverse transform unit 12 provides the frame buffer 14 with a local decoded image obtained by combining the orthogonal inversely transformed data and the reference image.

  The frame buffer 14 holds a frame of a locally decoded image. The frame buffer 14 gives the held image frame to the intra prediction unit 6 and the buffer management unit 15 as a reference image frame.

  The intra prediction unit 6 receives the reference image frame from the frame buffer 14, performs intra prediction between the current image frame and the reference image frame, and transmits the predicted image to the orthogonal transform unit 8 and the entropy encoding unit via the mode selection unit 13. 10 is given.

  The buffer management unit 15 manages reference image frames held in the frame buffer 14.

  The frame buffer 16 holds the reference image frame given from the buffer management unit 15, and gives the held reference image frame to the inter prediction unit 5 and the mv detection unit 4.

  The inter prediction unit 5 performs inter prediction based on the reference image frame from the frame buffer 16 and the motion vector from the mv detection unit 4. In addition, the inter prediction unit 5 gives the predicted image to the orthogonal transform unit 8 via the mode selection unit 13.

  The mode selection unit 13 selects an optimal one of the prediction images obtained by the inter prediction unit 5 and the intra prediction unit 6, and gives the selected prediction image to the orthogonal transform unit 8. The mode selection unit 13 notifies the selected prediction mode to the encoding control unit 3 and the buffer management unit 15.

  The object management unit 1 manages a target area occupied by a specified target specified in an image. When the object management unit 1 receives the input target area set, the object management unit 1 searches for and tracks a block that most closely matches (most closely resembles) a block included in the target area set of the reference image from the peripheral block set of the input image. To do. The object management unit 1 gives the search area control unit 2 a set of target areas to be recorded. Further, the object management unit 1 receives the motion vector from the mv detection unit 4, updates the block set of the target region based on the received motion vector, records the updated block set of the target region, and updates the block set. A block set of the target area is given to the encoding control unit 3.

  Here, the target area set is a set of areas to be noticed, and the object management unit 1 manages the target area as, for example, a block number of a block, a pixel position, or a designated range of each row of an image. Various methods can be applied to the method of specifying the target area set.

  The peripheral block set is a set of block numbers of blocks located in the periphery within a predetermined distance from the target area set of the input image. Further, the distance from the target area set for determining the peripheral block set is assumed to be larger than the normal search range.

  FIG. 2 is a block diagram illustrating a functional configuration of the object management unit 1. As illustrated in FIG. 2, the object management unit 1 includes a target area set recording unit 101, a peripheral block set calculation unit 102, and a target area set update unit. 103.

  The target area set recording unit 101 has a memory for recording the target area set and the target block set and an input / output control function. When the target area set is received, the target area set is recorded, and the target area set is recorded. This is given to the peripheral block set calculation unit 102. In addition, the target area set recording unit 101 records the target area set updated by the target area set update unit 103 according to the motion vector.

  When the peripheral block set calculation unit 102 receives the target region set from the target region set recording unit 101, the peripheral block set calculation unit 102 calculates a block number of a peripheral block located within a predetermined distance from the target region set and generates a peripheral block set. is there. The peripheral block set calculation unit 12 gives the obtained peripheral block set to the search range control unit 2.

  Upon receiving the motion vector from the mv detection unit 4, the target area set update unit 103 generates a target block set based on the motion vector, and uses the generated target block set as a new target area set. To be recorded.

  The search range control unit 2 controls the range in which the mv detection unit 4 detects a motion. When the search range control unit 2 receives a peripheral block set from the object management unit 1, the search range control unit 2 performs normal search range for blocks belonging to the peripheral block set. In addition, a part of the target area block calculated by the peripheral block set calculation unit 102 (within the search range used by the peripheral block set calculation unit 102) is searched, and the search range of other blocks is narrowed. is there.

  The mv detection unit 4 detects a motion vector by comparing the reference image frame from the frame buffer 16 with the input image frame from the block division unit 7. Further, the mv detection unit 4 gives the detected motion vector to the object management unit 1, the inter prediction unit 5, and the entropy encoding unit 10.

  When the encoding control unit 3 receives the target block set from the object management unit 1, the encoding control unit 3 controls the target block set to reduce the quantization step size. Control is performed to increase the quantization step size. In this way, by reducing the quantization step size for the target block set, it is possible to reduce the distortion of the DCT coefficient value when performing decoding processing on the decoding side, so that the designated target in the image is clearly displayed. be able to.

(A-2) Operation of the First Embodiment Next, the operation of the moving image encoding process of the image encoding device 100 of the first embodiment will be described.

  In FIG. 1, first, when a target area designation to be noticed is received, a target area set corresponding to the area designation is given to the object management unit 1.

  When the target area set is given to the object management unit 1, the target area set is recorded in the target area set recording unit 101 of the object management unit 1, and the target area set is given to the peripheral block set calculation unit 102, and the peripheral block The set calculation unit 102 calculates a neighboring block set.

  Here, FIG. 3 is a flowchart showing the peripheral block calculation processing in the peripheral block set calculation unit 102, and FIG. 4 is an explanatory diagram for explaining the relationship between the target area block and the peripheral block set.

  In FIG. 3, first, the inputted image is divided into blocks (S11). At this time, you may make it use the block divided | segmented by the block division part 7 shown in FIG.

  Next, when the target area set is received from the target area set recording unit 101, it is determined whether or not the target area set includes one corresponding to each block of the reference image, and the target area set is included in each block of the reference image. When included, blocks corresponding to the target area are identified (S12), and these blocks are set as target area blocks.

  Then, a block located within a predetermined range from each block constituting the target area block is specified as a peripheral block (S13), and block numbers including the target area block and the peripheral block are calculated as a peripheral block set.

  FIG. 4 shows an example in which a block located within two blocks from each block constituting the target area block is a peripheral block. Also, as shown in FIG. 4, it is assumed that the peripheral block set obtained by the peripheral block set calculation unit 102 includes the target area block.

  Thus, when the peripheral block set is calculated by the peripheral block set calculating unit 102, the peripheral block set is given to the search range control unit 2, and the search range control unit 2 controls the search range.

  That is, the search range control unit 2 controls the blocks constituting the peripheral block set to expand the search range by adding a part of the target area block to the normal search range, and blocks other than the blocks of the peripheral block set Control to narrow the search range.

  Thus, since the search range of the target area can be expanded, it is possible to deal with even if the target to be noticed has a wide area. On the other hand, although the calculation amount of motion search related to the target region increases, the calculation amount of motion search of the portion other than the target region can be reduced, so that the overall processing load is large. Instead, it can be intensively calculated for an object to be noticed to make it clear.

  When the search range controlled by the search range control unit 2 is given to the mv detection unit 4, the mv search unit 4 detects a motion vector based on the reference image frame and the input image frame using the search range, The detected motion vector is given to the object management unit 1.

  When the motion vector from the mv detection unit 4 is given to the object management unit 1, the target region set update unit 103 of the object management unit 1 updates the target vector to a new target region, and the new target region is recorded in the target region set record. Recorded in the unit 101.

  The contents of the processing in FIGS. 5 and 6 are processed based on the following steps S21 to S38.

<Step S21, FIG. 5>
First, when the position indicated by the motion vector of a block belongs to the target area set among all the blocks constituting the peripheral block set, the block belonging to the peripheral block set is put into the target block set.

<Step S22: FIG. 5>
Next, a connected component in the target block set is obtained. Here, a connected component is a set of blocks of adjacent blocks, and a block belongs to a connected component means that it can reach an arbitrary block in the connected component through a block belonging to the adjacent connected component. is there. Each block obtained as a connected component is assigned a block number in order of 1 to 2, 3,..., N, so that each connected component can be represented by a set of block numbers of blocks belonging to the connected component. Like that. FIG. 7 shows an example in which nine connected components exist. The connected component L (1) is a set composed of block numbers: 1 to 6, the connected component L (2) is a set composed of block numbers: 1 to 25, and the connected component L (3) is composed of block numbers: 1 to 5. The connected component L (4) is a set consisting of block numbers: 1 to 5, the connected component L (5) is a set consisting of block numbers: 1 to 25, and the connected component L (6) is a block number: 1-24. The connected component L (7) is a set consisting of block numbers: 1 to 5, the connected component L (8) is a set consisting of block numbers: 1 to 5, and the connected component L (g) is a block number 1-24. Is a set of

<Step S23: FIG. 5>
Next, the obtained number of connected components (9 in the example of FIG. 7) is substituted into the variable n.

<Step S24: FIG. 5>
If the size of the connection portion L (n) (for example, the number of blocks constituting the connection portion L (n)) is less than a certain size, the process proceeds to step S25, and if greater than a certain size, the process proceeds to step S26.

<Step S25: FIG. 5>
The connected component L (n) is excluded from the target block set.

<Step S26: FIG. 5>
Subtract 1 from variable n.

<Step S27: FIG. 5>
The process from step S24 to step S26 is repeated until the variable n becomes 0. When the variable n becomes 0, in all of the obtained connected components, the size of each connected component (the number of blocks constituting each connected portion) is less than a certain threshold value is excluded from the target block set. For example, the target block set is a connected component L (9) obtained from the connected component L (1) obtained as shown in FIG. In step S21 to step S27, if the size of the connected component (the number of blocks constituting the connected portion) is 6 or less, when the connected component L (n) is excluded from the target block set, the connected component L (1), connected Component L (3), connected component L (4), connected component L (7), connected component L (8) are excluded, connected component L (2), connected component L (5), connected component L (6) The connected component L (9) remains. FIG. 8 shows an example of a result obtained by performing the processing from step S23 to step S27 on FIG. The number of connected components in FIG.

<Step 528 to Step 536>
In each connected component, blocks whose motion vectors are significantly different from other motion vectors are excluded from the connected components (steps S28 to S36).

<Step S28: FIG. 5>
First, a representative value of motion vectors of each connected component in the target block set (for example, an average value of motion vectors of each connected component in the target block set) is obtained. FIG. 9 shows an example of the result obtained by performing the process of step S28 on FIG.

<Step S29: FIG. 5>
Next, the number of connected components obtained in step S22 (9 in the example of FIG. 9) is substituted for variable n.

<Step S30: FIG. 6>
The variable n is compared with the value n when the connected component L (n) is excluded in step S25. As a result of the comparison, if the variable n is the same as the value obtained when the connected component L (n) is excluded in step S25: n (n = 1, 3, 4, 7, 8 in FIG. 9), step S31 is performed. Proceed to As a result of the comparison, if the variable n is different from the value obtained when the connected component L (n) is excluded in step S25: n (n = 1, 3, 4, 7, 8 in FIG. 9), step S32 (FIG. 9 )

<Step S31: FIG. 6>
1 is subtracted from the variable n, and the process returns to step S30.

<Step 532: FIG. 6>
The variable m is assigned the size of the connected component L (n) (here, for example, the number of blocks M (n) constituting the connected component L (n)). In FIG. 9, M (2) = 25, M (5) = 25, M (6) = 24, and M (9) = 24.

<Step 533: FIG. 6>
The absolute value (ABS) of the value obtained by subtracting the magnitude of the representative value of the motion vector in the connected component L (n) from the magnitude of the motion vector in the connected component L (n) and block number m is compared with a predetermined threshold value. To do. If the obtained absolute value (ABS) is larger than the predetermined threshold value, the process proceeds to step S34, and if the obtained absolute value (ABS) is less than the predetermined threshold value, the process proceeds to step S35.

<Step 534: FIG. 6>
When the absolute value (ABS) obtained in step S33 is larger than a predetermined threshold, the block m is excluded from the connected component L (n).

  In FIG. 9, in the connected component L (2), when the representative value of the motion vector is the motion vector V2, the motion vector of the block number m = 1: v (2,1) [V21] and the block number m = 3 Motion vector: v (2, 3) [V22], motion vector with block number m = 12: v (2, 12) [V23], motion vector with block number m = 17: v (2, 17) [V24] In the motion vector of block number m = 24: v (2, 24) [V25], the motion vector of block number m = 25: v (2, 25) [V26], the absolute value (ABS) obtained in step S33. ) Is greater than a predetermined threshold. Therefore, as shown in FIG. 10, in the connected component L (2), the block with the block number m = 1: b (2, 1) [B21], the block with the block number m = 3: b (2, 3) [ B22], block with block number m = 12: b (2,12) [B23], block with block number m = 17: b (2,17) [B24], block with block number m = 24: b (2 24) [B25], block number m = 25: b (2.25) [B26] is excluded from the connected components.

In FIG. 9, in the connected component L (5), when the representative value of the motion vector is the motion vector V5, the motion vector of the block number m = 11: v (5,11) [V51], the block number m = 16 Motion vector: v (5, 16) [V52], motion vector with block number m = 24: v (5, 24) [V53], motion vector with block number m = 25: v (5, 25)
In [V54], the absolute value (ABS) obtained in step S33 is greater than a predetermined threshold value. Therefore, as shown in FIG. 10, in the connected component L (5), the block of block number m = 11: b (5, 11) [B51], the block of block number m = 16: b (5, 16) [ B52], block with block number m = 24: b (5, 24) [B53], block with block number m = 25: b (5, 25) [B54] are excluded from the connected components.

  In FIG. 9, in the connected component L (6), when the representative value of the motion vector is the motion vector V6, the motion vector of the block number m = 1: v (6,1) [V61] and the block number m = 24 In the motion vector: v (6, 24) [V62], the absolute value (ABS) obtained in step S33 is larger than a predetermined threshold value. Therefore, as shown in FIG. 10, in the connected component L (6), the block with block number m = 1: B (6, 1) [B61], the block with block number m = 24: B (6, 24) [ B62] is excluded from the connected components.

  In FIG. 9, in the connected component L (9), when the representative value of the motion vector is the motion vector V9, the motion vector of the block number m = 1: v (9,1) [V91], the block number m = 3 Motion vector: v (9,3) [V92], motion vector with block number m = 6: v (9,6) [V93], motion vector with block number m = 11: v (9,11) [V94] In step S33, the absolute value (ABS) obtained in step S33 is larger than a predetermined threshold value. Therefore, as shown in FIG. 10, in the connected component L (9), the block with block number m = 1: b (9, 1) [B91], the block with block number m = 3: b (9, 3) [ B92], block with block number m = 6: b (9,6) [B93] and block b (9,11) [B94] with block number m = 11 are excluded from the connected components.

<Step S35: FIG. 6>
Subtract 1 from the variable m.

<Step S36: FIG. 6>
Until the variable m becomes 0, the processing of step S33 to step S35 is repeated. When the variable m becomes 0, in the connected component L (n), when the absolute value (ABS) obtained in step S33 is larger than a predetermined threshold, the process of excluding the block m from the connected component L (n) is completed. .

<Step S37: FIG. 6>
Subtract 1 from variable n.

<Step S38: FIG. 6>
Steps S30 to S37 are repeated until the variable n becomes zero. When the variable n becomes 0, in each connected component, when the absolute value (ABS) obtained in step S33 is larger than a predetermined threshold, the process of excluding the block m from each connected component is completed. An example of the result of processing the target block set shown in FIG. 7 based on steps S21 to S38 is shown in FIG.

  In this way, a region corresponding to the generated target block set is set as a new target region set, and the object management unit 1 gives this target block set to the encoding control unit 3.

  The encoding control unit 3 receives the target block set from the object management unit 1 and controls the target block set so as to reduce the quantization step size. Control to increase the step size.

  As described above, since the encoding method for the input image other than the search range and quantization step size control for the target region can be applied, the detailed description here is omitted. .

  Next, an operation when the matching pattern unit 200 is provided in the preceding stage of the image encoding device 100 according to the first embodiment will be described with reference to FIG.

  As shown in FIG. 12, a pattern matching unit 200 is connected to the previous stage of the image encoding device 100 of the first embodiment.

  When an input image is input to the pattern matching unit 200, the input image is compared with a pattern prepared in advance in the pattern matching unit 200, and the matched region is converted into a target region set and given to the image encoding device 100. .

  When the target region set is input to the image encoding device 100, the image encoding device 100 generates a peripheral block set using the target region set, and based on the peripheral block set, motion search for blocks belonging to the peripheral block set Control the motion search to expand the range and narrow the search range of parts other than the surrounding block set, update the target region set using the obtained motion vector, and use the corresponding target block set to make the smaller quantization to the target block Encoding control is performed so as to allocate the step size.

(A-3) Effects of the First Embodiment As described above, according to the first embodiment, even if the target motion is large by expanding the motion search range of the peripheral blocks in the target region, Tracking can be performed, and the target for reducing the quantization step size can be ensured.

  Further, according to the first embodiment, the calculation amount can be kept constant by narrowing the search range of other blocks, and can be executed even with hardware having relatively low performance.

Furthermore, according to the first embodiment, by appropriately allocating the code amount, it is possible to suppress the code amount while clearly encoding a target of interest even when the overall code amount is limited. (B) Second Embodiment Next, a second embodiment of the image coding apparatus of the present invention will be described with reference to the drawings.

  The difference between the image coding apparatus of the second embodiment and the image coding apparatus of the first embodiment is that, in the object management unit, each connected component and each connected component of the target block set calculated by the target region set update unit. The representative value of the motion vector is recorded.

  Therefore, hereinafter, the function of the image encoding device according to the second embodiment and an example in which the image encoding device according to the second embodiment is applied as an image encoding device of a surveillance video system will be described.

(B-1) Configuration and Operation of Second Embodiment FIG. 13 is a configuration diagram showing an application example when the image encoding device 110 of the second embodiment is applied to a surveillance video system.

  As illustrated in FIG. 13, the image encoding device 110 according to the second embodiment is connected to a pattern matching unit 200 and a motion analysis unit 300.

  Similar to the first embodiment, the pattern matching unit 200 compares an input image with a pattern prepared in advance in the pattern matching unit 200, converts the matched region into a target region set, and performs image coding. It is given to the device 100.

  When receiving the target region set from the pattern matching unit 200, the image encoding device 110 performs motion search control and code amount control, as in the first embodiment. In addition, the image encoding device 110 according to the second embodiment provides the motion analysis unit 300 with the connected components constituting the target block set as the target region and the representative value of the motion vector.

  Note that the internal configuration of the image encoding device 110 of the second embodiment corresponds to the internal configuration of the first embodiment shown in FIG. 1, but the function of the object management unit is different.

  FIG. 14 is a block diagram illustrating a functional configuration of the object management unit according to the second embodiment. The object management unit according to the second embodiment illustrated in FIG. 14 includes a target area set recording unit 104 and a peripheral block set calculation unit. 102, and a target area set update unit 103.

  Since the peripheral block set calculation unit 102 and the target region set update unit 103 correspond to the configuration described in the first embodiment, detailed description thereof is omitted here.

  Similar to the first embodiment, the target area set recording unit 104 includes a memory for recording the target area set and an input / output control function, and each of the connected blocks constituting the target block set calculated by the target area set update unit 103. A memory for recording a representative value of the component and a motion vector of each connected component; The target area set area 104 outputs each connected component constituting the target block set and a representative value of a motion vector of each connected component as movement information.

  When the motion analysis unit 300 receives a connected component constituting the target block set and a representative value of the motion vector from the image encoding device 110, the motion analysis unit 300 performs motion tracking of the target region based on the connected component and the representative value of the motion vector. Is to do. Also, the motion analysis unit 300 determines that the tracking target is within the frame of the input image, and when trying to move out of the frame, the motion analysis unit 300 performs camera control so that the tracking target is within the frame of the input image. (Not shown). Furthermore, the motion analysis unit 300 records motion analysis log information in a recording unit (not shown) of target motion analysis information.

(B-2) Effects of the Second Embodiment As described above, according to the second embodiment, the same effects as those described in the first embodiment can be achieved.

  In addition, according to the second embodiment, since a function for outputting motion information is provided, it is not necessary to provide a separate tracking function, and therefore, it can be executed even by an apparatus using hardware with low performance.

(C) Third Embodiment Next, a third embodiment of the image encoding device of the present invention will be described with reference to the drawings.

  The image coding apparatus according to the third embodiment is different from the image coding apparatus according to the first embodiment in that the object management unit controls changes such as addition and deletion of a target area set.

  Therefore, hereinafter, the function of the image encoding device of the third embodiment and an application example of the image encoding device of the third embodiment will be described.

(C-1) Configuration and Operation of Third Embodiment FIG. 15 is a configuration diagram illustrating an application example of the image encoding device 120 of the second embodiment. In FIG. 15, the image encoding device 120 is a target. The area designation unit 400 is connected.

  The target area designating unit 400 receives, for example, an operation of a user or the like, takes in the area designation of the tracking target, and gives the area designation to the image encoding device 120. In addition, the target area designating unit 400 accepts designation of a new area to be tracked, and captures designation changes such as deletion of an already designated target.

  As a result, the motion search range is expanded only for the region specified by the target region specifying unit 400, and the quantization step size can be greatly reduced for other portions.

  Note that the target area specifying unit 400 can be applied, for example, to capture a target specified by an operation of a touch panel, a pointing device, or the like.

  Similar to the first embodiment, the image encoding device 120 performs motion search control and code amount control, and receives a control signal for specifying a target from the target region specifying unit 400, and the specified target This is for adding or deleting area sets.

  FIG. 16 is a block diagram illustrating a functional configuration of the object management unit according to the third embodiment. The object management unit according to the second embodiment illustrated in FIG. 16 includes a target area set recording unit 105 and a peripheral block set calculation unit. 102, and a target area set update unit 103.

  Since the peripheral block set calculation unit 102 and the target region set update unit 103 correspond to the configuration described in the first embodiment, detailed description thereof is omitted here.

  Similar to the first embodiment, the target area set recording unit 105 includes a memory for recording the target area set and an input / output control function, and receives a control signal for specifying a target from the target area specifying unit 400. The specified target area set is added or deleted.

(C-2) Effects of Third Embodiment As described above, according to the third embodiment, the same effects as those described in the first embodiment can be achieved.

  In addition, according to the third embodiment, by including the target area control function, it can be included in the tracking target interactively or excluded. As a result, this is effective when there is an object that is clearly projected and an object that is desired to be blurred.

(D) Other Embodiments The first to third embodiments described above can be widely applied to an encoding apparatus that encodes a moving image, but can be applied to, for example, an image encoding apparatus of a surveillance video system. When applied to a surveillance video system, it is possible to allocate a large amount of calculation and code necessary for encoding a specific designated object, so that the object of the designated area can be obtained as a clear image, Tracking and motion analysis are also easy. In addition, an application may be considered in which image processing is automatically performed in real time so that only a specific object is sharpened and the other portions are blurred by increasing the degree of allocation of the calculation amount and code amount using this apparatus.

  It is also possible to provide an image encoding device that has both the functions of the image encoding devices described in the second and third embodiments.

  The functions of the image coding apparatus described in the first to third embodiments described above can be realized by hardware processing such as a CPU executing software processing, but are realized as hardware. May be.

It is a block diagram which shows the internal structure of the image coding apparatus of 1st Embodiment. It is a block diagram which shows the function structure of the object management part of 1st Embodiment. It is a flowchart which shows the calculation process of the surrounding block set of 1st Embodiment. It is explanatory drawing explaining the relationship between the object area | region block of 1st Embodiment, and a periphery block set. It is a flowchart (sparse 1) which shows the calculation process of the object block set of 1st Embodiment. It is a flowchart (the 2) which shows the calculation process of the object block set of 1st Embodiment. It is a figure which shows an example of the block which belongs to the connection part in the object block set of 1st Embodiment. It is a figure which shows an example of the block which belongs to the connection part in the object block set processed based on FIG.7 S23-S27. It is a figure which shows an example of the block which belongs to the connection part in the object block set processed based on FIG.8 with respect to FIG. It is a figure which shows an example of the block which belongs to the connection part in the object block set processed based on FIG.8 S29-S34. It is a figure which shows an example of the block which belongs to the connection part in the object block set processed based on FIG.7 S21-S38. It is a block diagram which shows the example of application of the image coding apparatus of 1st Embodiment. It is a block diagram which shows the example of application of the image coding apparatus of 2nd Embodiment. It is a block diagram which shows the function structure of the object management part of 2nd Embodiment. It is a block diagram which shows the example of application of the image coding apparatus of 3rd Embodiment. It is a block diagram which shows the function structure of the object management part of 3rd Embodiment.

Explanation of symbols

  DESCRIPTION OF SYMBOLS 1 ... Object management part, 101, 104 and 105 ... Target area set recording part, 102 ... Peripheral block set calculation part, 103 ... Target area set update part, 2 ... Search range control part, 3 ... Encoding control part, 4 ... mv detection unit, 5 ... inter prediction unit, 6 ... intra prediction unit, 7 ... block division unit, 8 ... orthogonal transform unit, 9 ... quantization unit, 10 ... entropy coding unit, 11 ... inverse quantization unit, 12 ... Orthogonal inverse transform unit, 13 ... mode selection unit, 14 ... frame buffer, 15 ... buffer management unit, 16 ... frame buffer, 100, 110 and 120 ... image encoding device, 200 ... pattern matching unit, 300 ... motion analysis unit, 400... Target area designation unit.

Claims (4)

In an image encoding device for encoding a moving image,
Target area management means for managing a target area indicating a position of a specified target in a specified image;
A peripheral region determining means for determining a peripheral region including a target region to be specified in the image;
A search range in which the search range for motion search is expanded for the peripheral region determined by the peripheral region determination means, and the search range for motion search is narrowed for regions other than the peripheral region. Control means;
Motion vector detection means for detecting a motion vector within the search range controlled by the search range control means;
Target area update means for updating the target area to be specified in the image based on the motion vector detected by the motion vector detection means and the peripheral area;
An image encoding apparatus comprising: an encoding control unit configured to encode a target region to be specified in the image managed by the target region management unit with a quantization step size set small.   In the new target region updated by the target region update means, the motion of each connected component based on one or a plurality of connected components configured by grouping adjacent blocks and the motion vector of each block constituting each connected component 2. The image encoding apparatus according to claim 1, wherein a representative value of the vector is obtained.   The target area management means holds a new target area updated by the target area update means, each connected component in the target area, and a representative value of a motion vector of each connected component. Item 3. The image encoding device according to Item 2. In an image encoding method for encoding a moving image,
A target area management means for managing a target area indicating a position of a specified target in a specified image;
A peripheral region determining step for determining a peripheral region including a target region to be specified in the image;
The search range control means widens the search range of motion search for the peripheral area determined by the peripheral area determination means, and narrows the search range of motion search for areas other than the peripheral area. A search range control step for controlling
A motion vector detection step in which the motion vector detection means detects a motion vector within the search range controlled by the search range control means;
A target region update step in which the target region update unit updates the target region of the designation target in the image based on the motion vector detected by the motion vector detection unit and the peripheral region;
An encoding control unit comprising: an encoding control step for encoding with a small quantization step size set for a target region to be specified in the image managed by the target region management unit. An image encoding method to be performed.

Images