JP2002077905A - Encoding controller and controlling method - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide an encoding controller and controlling method realizing an encoded image transmission using the transmission band efficiently by enhancing the accuracy of encoding control and minimizing the margin of a target generated information quantity. SOLUTION: A frame unit quantization control section 111 determines quantization characteristics being used in a current image frame based on the generated information quantity of the preceding image frame. A block unit generated information quantity calculating section 112 calculates generated information quantity per block from generated information quantity accumulated up to the current time point for the current image frame. A maximum allowable information quantity calculating section 114 calculates maximum allowable information quantity per block every time when one block in the current image frame is encoded. A comparing section 113 compares generated information quantity per block with a maximum allowable information quantity every time when one block in the current image frame is encoded. Based on the comparison results, a block unit quantization control section 115 determines quantization characteristics of next block within a predetermined range around quantization characteristics determined at the frame unit quantization control section 111.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an encoding control apparatus and method, and more particularly, to an encoding control apparatus for compressing and encoding a signal such as a video signal with high efficiency and transmitting the same, and the apparatus. The present invention relates to an encoding control method.

[0002]

2. Description of the Related Art A technique for efficiently compressing and transmitting a video signal and transmitting the signal is generally applied to the fields of videophone and videoconference. In the future, this technology will be applied to surveillance systems that monitor dangerous locations and transmit video between mobile objects by transmitting video signals by digital radio using a wireless LAN transmission path, and to use video over the Internet. Application to distribution is expected. Hereinafter, a conventional image encoding method will be described in detail using an example of image encoding performed in a video conference or a videophone.

Conventionally, in the field of videoconferencing and videophone, inter-frame predictive coding using temporal correlation between frames and intra-frame predictive coding using spatial correlation within a frame are known. Combined coding is commonly used. A television image composed of 30 images (frames) per second has a large correlation in the time axis direction. Therefore, if the pixel at the same position in the screen one image frame before is used for the prediction of the current image frame using the inter-frame correlation, the most ideal prediction can be performed when the screen is stationary. Become. However, in the inter-frame predictive coding, the correlation between frames is rather low in the case of a video with many scene changes or a video with a large movement of a subject on a screen (compared to the correlation between adjacent pixels in a field). Is also lower).

On the other hand, each pixel of the image signal for each frame is also
The level change between adjacent pixels is small and the correlation is strong. It is said that the autocorrelation function can be approximated by a negative exponential function. At this time, the power spectrum density, which is the Fourier transform of the autocorrelation function, has the property of being maximum at the zero frequency component (that is, the DC component) and monotonically decreasing as the frequency component increases. The best known orthogonal transform to the frequency domain is the Fourier transform.
However, since the Fourier transform involves a complex number operation and has a complicated configuration, in image coding, a two-dimensional DCT (Discrete Cosine Transform) is used as an alternative orthogonal transform.
Is generally used. Transform coefficients (image data) decomposed into frequency components by DCT include level zero, which is a transform coefficient (insignificant coefficient) that is not coded, and level ± 1, which is a significant coefficient that takes a discrete quantized representative value, from level ± 1 to level ±.
K (K is an arbitrary integer) is quantized. afterwards,
The quantized transform coefficients are compressed by performing run-length coding for coding the continuity of insignificant coefficients and Huffman coding for assigning variable-length codes according to the occurrence probability of the level of the significant coefficient. .

For example, an ITU (International Telecommu)
H. nication Union) -T. In H.261, motion-compensated inter-frame prediction is applied to an image with small motion, and the following encoding is performed on the prediction error between frames. In addition, for an image with large motion, the following encoding is directly performed on the frame pixels without applying the motion compensated inter-frame prediction. FIG. 7 shows ITU-
Recommendation T. H. 261 shows an encoder and a decoder for image data. As shown in FIG. 2
The image data encoder 700 of FIG.
1, an orthogonal transform unit 702 that performs two-dimensional DCT, a quantization unit 703, an inverse quantization unit 704, and an inverse orthogonal transform unit 705.
, An adder 706, an image memory 707 having a motion compensation function, a filter 708 in a loop, and an encoding controller 7.
09, and selectors 710 and 711.
On the other hand, the decoder 750 includes an inverse quantization unit 751, an inverse orthogonal transformation unit 752, an addition unit 753, an image memory 754 having a motion compensation function, an in-loop filter 755, and a selector 756. .

In the encoder 700, a subtractor 701
Is a 352 x 288 dot CIF (Common Interm
By calculating a difference between the video input signal converted into the edit format (edit format) and the prediction data stored in the image memory 707, a prediction error between frames is calculated. At this time, motion compensation is performed in the image memory 707. Specifically, the image memory 707 most closely resembles the encoding target block within a range of ± 15 pixels around each pixel of the block having the same block address as the encoding target block (the difference error is small). An area of 16 × 16 pixels is detected from the reference frame memory, and the motion amount is output as a two-dimensional motion vector. This two-dimensional motion vector is
The data is sent to the decoder 750 together with the image data (transform coefficients). For large motions for which motion compensation is not effective, selectors 710 and 711 select intra-frame predictive coding for which no prediction is performed. Orthogonal transform unit 7
In step 02, prediction errors and frame pixels are divided into blocks of 8 × 8 pixels, and two-dimensional DCT is performed on each block. With this two-dimensional DCT, the pixels of each block are
Converted to frequency components. The transform coefficient obtained by the transform is quantized by the quantization unit 703. By this quantization, each transform coefficient is represented by a level of a significant coefficient which is an integer from the level of the insignificant coefficient ± 1 to the level ± K. The quantized transform coefficients are sent to a decoder 750 via a communication unit (not shown) and the like,
04 and inversely transformed by the inverse orthogonal transform unit 705, the current predicted data (image memory 707)
, And stored in the image memory 707 as prediction data to be used for the next encoding.

On the other hand, in the decoder 750, the input transform coefficients (image data) are inversely transformed by the inverse quantization unit 751 and the inverse orthogonal transform unit 752, and then stored in the image memory 754 by the adder 753. Is added to the existing prediction data. The added data is a video output and is stored in the image memory 754 as the next prediction data. When the input data is intra-frame encoded data, the prediction data is not selected in the selector 756, and the input data is directly inversely converted, taken out as a video output, and stored in the image memory 754. . The above is an example of predictive coding of a video signal, in particular, coding in which inter-frame predictive coding and intra-frame predictive coding are combined.

Next, a conventional encoding control method in the encoding control unit 709 will be described. Since the spatial correlation and the temporal correlation of an image change every moment, the information amount as a result of removing the redundancy by the DCT or the inter-frame difference also changes every moment. The amount of information after encoding is changed by changing a quantization characteristic applied to information obtained by performing DCT.
Can be controlled. However, compression of information by quantization involves image distortion due to quantization errors. Coding control is achieved by selecting quantization characteristics to bring the amount of generated information closer to a desired value. Therefore, when the frame rate is kept constant and the transmission rate is constant, the redundancy that fluctuates according to the characteristics of the input image is predicted, and the quantization characteristics are selected so that the amount of generated information is always uniform. Is done.

For example, in a report "Study of Low-Delay Interframe Prediction Coding" published by Senda et al. In the IEICE Autumn Conference D-162, 1992, an image frame is called a 16 × 16 pixel called a macroblock. , And based on the difference between the information amount generated up to the current macroblock and the prediction value, determine the quantization characteristic of the next macroblock to be encoded each time the macroblock is encoded. Is disclosed. In this report, the quantization information (quantization parameter) of the next macroblock is determined on the basis of the target information generation amount in macroblock units and the actually generated information amount, so that the information generation amount of one image frame is determined. Is controlled to approach the target amount of generated information of one image frame. If the target generated information amount in units of macroblocks is equalized within a frame, the above-described control processing is simple. However, if the same target generated information amount is used for a macroblock having a high inter-frame correlation and a macroblock which is not so, the motion There is a problem that the image quality of a large block is deteriorated.
In most cases, a moving block is often the target of interest, and the image quality of the target of interest deteriorates.

On the other hand, Japanese Patent Laid-Open Publication No. Hei 5-344493 discloses a moving picture coding apparatus which solves the above problem. In this publication, an encoding parameter is determined based on an output from a predictive coding efficiency calculator that calculates an interframe predictive coding efficiency, so that optimal coding is performed for each moving image portion. A method is disclosed. However, even with this method, it is not easy to determine the target generated information amount for each macroblock according to the correlation within the frame while keeping the generated information amount of one image frame within the target generated information amount. It involves complicated processing.

Therefore, it is common to determine the quantization characteristic by referring to a predetermined encoding parameter control table based on the amount of information generated or the buffer occupation amount of the encoding result one image frame before. Is being done. However, since the relationship between the amount of generated information and the quantization characteristic is not uniquely determined, for example, Japanese Unexamined Patent Publication No. 4-57489.
As disclosed in Japanese Unexamined Patent Application Publication, it is considered to have an encoding parameter control table for each of assumed scene types. Here, the buffer occupancy is the occupancy of the image data in the transmission buffer for temporarily storing the coded image data for transmission.

Further, the applicant of the present application has disclosed in
Japanese Patent Publication No. 482 proposes the following determination method instead of uniquely determining the quantization characteristic based on the amount of generated information or the buffer occupancy of the encoding result one image frame before. In this method, the upper limit and the lower limit of the target generated information amount for the quantization characteristic are set in advance. If the generated information amount resulting from encoding one image frame with a certain quantization characteristic is between the upper limit and the lower limit of the target generated information amount, the quantization characteristic using the same quantization characteristic for the next image frame is used. If the upper limit of the target generated information amount is exceeded, the quantization characteristic whose quantization precision is reduced from the current state (that is, the image quality is degraded) is determined as the quantization characteristic used for the next image frame. If the amount of generated information is lower than the lower limit, the quantization characteristic with improved quantization accuracy (that is, improved image quality) is determined as the quantization characteristic used for the next image frame. According to this encoding control method, it is possible to adapt to any scene with simple control without specifying the absolute level. The operation will be described below with reference to FIG.

FIG. 8 is a diagram showing a temporal transition of the amount of information generated as a result of encoding. The vertical axis in FIG. 8 indicates the amount of generated information, and the upward direction of the vertical axis is the direction in which the amount of generated information increases. The horizontal axis in FIG. 8 indicates the quantization characteristic (Quant), and the magnitude of the quantization characteristic corresponds to the magnitude of the quantization step size (quantization error). In the direction in which the conversion step size increases. When an image of 30 frames / second is always transmitted at a constant transmission speed, the amount of information generated per image frame is ideally set to the transmission speed / 30 frames. However, in reality, it is impossible to accurately control the amount of generated information, so that the range of the generated information amount with an appropriate margin is set as the target generated information amount. Here, the quantization characteristic is represented by q (s), and each quantization characteristic is identified by an index s which is a natural number. The quantization step size of the quantization characteristic q (s) is large for the index s. It is assumed to be larger. At this time, the state of the quantization characteristic q (s) of the image frame to be encoded is specified by the value of the index s.

In the coding control method proposed by the applicant of the present invention, an upper threshold value and a lower threshold value of the amount of generated information are determined for each state of the quantization characteristic. These are determined so as to sandwich the target generated information amount. The amount of information generated as a result of encoding a certain image frame with the quantization characteristic q (s) is
When the threshold value is not exceeded (the amount of generated information a), this state is maintained, and the encoding of the next image frame is also performed using the quantization characteristic q (s). A certain image frame is quantized q
When the amount of information generated as a result of the encoding in (s) exceeds the upper limit threshold (the amount of generated information b), the state transits to a state having a large quantization characteristic (a state having a large quantization step size),
The next image frame is encoded with the quantization characteristic q (s + 1). As a result, the next generated information amount becomes c. When the amount of generated information as a result of encoding a certain image frame with the quantization characteristic q (s) exceeds the lower threshold (the amount of generated information d),
The state transits to a state in which the quantization characteristic is small (a state in which the quantization step size is small), and as a result of encoding the next image frame with the quantization characteristic q (s-1), the next generated information amount becomes e.

[0015]

However, the above-mentioned conventional encoding control method for controlling the quantization characteristic for each image frame has the following problems. The quantization characteristic is
It is a set of predetermined discrete quantized representative values. 1
When image frames are quantized collectively by one quantization characteristic, errors are accumulated by the number of macroblocks constituting one screen, resulting in a difference in information amount. Therefore, it is assumed that the quantization characteristic is most finely controlled. The information amount also fluctuates greatly. For this reason, the control involves a large error. In particular, the encoding control method proposed by the present applicant, ie,
In the case of using the method of determining the increase / decrease of the quantization characteristic based on the amount of information generated or the buffer occupation amount of the encoding result one image frame before, even if the correlation of the images is constant. , The amount of generated information is a quantization characteristic q
(S) below the lower limit of the target generated information amount, and the quantization characteristic q (s-1) exceeds the upper limit of the target generated information amount.
When the quantization characteristic is q (s) and q (s-
A situation may occur in which the amount of generated information fluctuates between 1) and 2). That is, as shown in FIG. 8, a situation may occur in which the state of the quantization characteristic fluctuates between the generated information amount e ′ and the generated information amount d. If the amount of generated information does not exceed the target amount of generated information even when the amount of generated information is large, it is necessary to set the upper limit of this vibration to be smaller than the target amount of generated information. And the transmission band cannot be used efficiently.

SUMMARY OF THE INVENTION It is therefore an object of the present invention to improve the accuracy of encoding control, minimize the margin of the target generated information amount and use the transmission band efficiently, and perform encoding capable of encoded image transmission. A control device and method are provided.

[0017]

Means for Solving the Problems and Effects of the Invention In order to achieve the above object, the present invention has the following features. A first invention is an encoding control device that controls encoding of an image frame divided into a plurality of blocks by setting a quantization characteristic for each of the blocks, and includes an image frame (current image) to be processed. The reference quantization characteristic used for encoding the current image frame is determined based on the amount of generated information which is the amount of data generated by encoding the immediately preceding image frame (previous image frame) with respect to the current image frame. A frame-unit quantization control unit to be determined, a generated information amount calculation unit that calculates the generated information amount per block each time one of the blocks constituting the current image frame is encoded, A comparison unit that compares the threshold value and the amount of generated information per block calculated by the generated information amount calculation unit, based on a comparison result of the comparison unit and a reference quantization characteristic. And a block unit quantization control unit for setting the quantization characteristic used for the encoding of the next block of blocks.

As described above, according to the first aspect, the quantization characteristic of each block is set near the quantization characteristic determined by the frame unit quantization control unit. As a result, an image frame can be encoded using a quantization characteristic whose quantization value can be changed in a minimum unit (that is, “1”) or less. You can get closer. Therefore, it is possible to improve the accuracy of the encoding control, reduce the margin of the target generated information amount as much as possible, and perform the encoded image transmission using the transmission band efficiently.

According to a second aspect of the present invention, there is provided an encoding control device for controlling encoding of an image frame divided into a plurality of blocks by setting a quantization characteristic for each of the blocks. The reference quantization characteristic used for encoding the (current image frame) is set to the immediately preceding image frame (previous image frame) with respect to the current image frame.
A frame unit quantization control unit that determines based on the data occupancy (buffer occupancy) in the smoothing buffer used for data transmission after encoding, and one of the blocks that constitute the current image frame are encoded. Every time
A generation information amount calculation unit for calculating a buffer occupancy amount per block; a comparison unit for comparing a predetermined threshold value with a buffer occupancy amount per block calculated by the generation information amount calculation unit; A block-based quantization control unit that sets a quantization characteristic used for encoding a block next to the encoded block based on a comparison result of the units and a reference quantization characteristic.

As described above, according to the second aspect, the quantization characteristic of each block is set near the quantization characteristic determined by the frame unit quantization control unit. As a result, an image frame can be encoded using a quantization characteristic whose quantization value can be changed in a minimum unit (that is, “1”) or less, and the buffer occupancy of one image frame can be changed to the target buffer occupancy. You can get closer. Therefore, it is possible to improve the accuracy of the encoding control, reduce the margin of the target buffer occupancy as much as possible, and use the transmission band efficiently to perform the encoded image transmission. Also, since the buffer occupancy is used,
Encoding control can be performed in consideration of a change in the effective transmission rate (throughput) due to retransmission or the like.

[0021] The third and fourth inventions are directed to the first and second inventions.
The predetermined threshold value is a maximum allowable information amount per block, and a predetermined target generated information amount or a target buffer occupancy amount and a block constituting one image frame. And is calculated from the total number. As described above, according to the third and fourth aspects, the threshold can be set by a simple algorithm.

[0022] The fifth and sixth inventions are directed to the first and second inventions.
Wherein the predetermined threshold value is a maximum allowable information amount per block, and a predetermined target generated information amount or a target buffer occupation amount and an effective block ( It is calculated from the total number of blocks in which a coding coefficient has occurred when the image frame was coded. As mentioned above, the fifth
According to the sixth aspect, the maximum allowable information amount per block is calculated by using an effective block in an image frame, so that an accurate threshold value in consideration of an operation area of a subject can be calculated by a simple algorithm. Can be set.

A seventh invention is an invention according to the third and fifth inventions, wherein the target generated information amount is corrected according to an effective transmission rate (throughput). As described above, according to the seventh aspect, even when a transmission error occurs due to deterioration of the radio wave state and the effective transmission speed changes, it is possible to perform optimal coding control according to the radio wave state.

An eighth invention is an invention according to the seventh invention, wherein the correction of the target generated information amount is (effective transmission rate /
(Standard transmission speed). As described above, according to the eighth aspect, it is possible to correct the target generated information amount according to the effective transmission rate by a simple calculation.

A ninth invention is an invention according to the third and fifth inventions, wherein the target generated information amount is corrected in accordance with the state of the start or end of the movement of the subject included in the image. Features. As described above, according to the ninth aspect, it is possible to perform optimal encoding control according to the amount of movement of a subject.

The tenth and eleventh inventions are dependent on the fifth and sixth inventions, wherein the total number of valid blocks is an average of the total number of valid blocks in a plurality of past image frames. The encoding control device according to claim 5. As described above, according to the tenth and eleventh aspects, by setting the total number of effective blocks to the average of a plurality of past image frames, the maximum allowable information amount per block that is not affected by the temporary movement of the subject is obtained. Can be calculated.

A twelfth invention is a invention according to the first and second inventions, wherein the block unit quantization control section comprises:
Based on the comparison result of the comparison unit, determine the quantization characteristic of the next block to be encoded to be equal to, larger than, or smaller than the reference quantization characteristic. It is characterized by.

A thirteenth invention is an invention according to the first and second inventions, wherein the block unit quantization control section comprises:
The quantization characteristic of the block to be encoded first in the image frame is used as a reference quantization characteristic. The same as or larger than the block of
Alternatively, the quantization characteristic is determined to be one of smaller quantization characteristics.

A fourteenth invention is an invention according to the first and second inventions, wherein the block unit quantization control section comprises:
It is characterized in that a quantization characteristic of a block to be encoded next is determined within a preset width centered on a reference quantization characteristic.

As described above, according to the twelfth to fourteenth aspects, it is possible to set the quantization characteristic for each block to a value near the quantization characteristic set by the frame-based quantization control unit using a simple algorithm. it can. Therefore, there is no variation in quantization characteristics (uneven image quality) for each block in the image frame, and the amount of generated information of the image frame can be close to the target generated information amount or the target buffer occupation amount.

The fifteenth and sixteenth inventions are dependent on the third and fourth inventions, wherein the maximum allowable information amount is updated every time a block is coded according to the following equation. And Max_info = (target_info-total_info) / (total_BK-
code_BK) Max_info: Maximum allowable information amount target_info: Target generated information amount or target buffer occupancy total_info: Generated information amount or buffer occupation amount up to the present time in the current image frame totak_BK: Total number of blocks constituting one image frame codec_BK: Current image frame Number of blocks coded to date

The seventeenth and eighteenth inventions are inventions dependent on the fifth and sixth inventions, wherein the maximum allowable information amount is updated every time a block is encoded according to the following equation. And Max_info = (target_info-total_info) / (pre_active
_BK-active_BK) Max_info: Maximum allowable information amount target_info: Target generated information amount or target buffer occupation amount total_info: Generated information amount or buffer occupation amount in the current image frame up to the present time pre_active_BK: Total number of active blocks in the previous image frame active_BK: Current image Number of valid blocks so far in the frame

The nineteenth and twentieth inventions are dependent on the tenth and eleventh inventions, wherein the maximum allowable information amount is updated every time a block is coded according to the following equation. And Max_info = (target_info-total_info) / (ave_active
_BK-active_BK) Max_info: Maximum allowable information amount target_info: Target generated information amount or target buffer occupation amount total_info: Generated information amount or buffer occupation amount up to the present time in the current image frame ave_active_BK: Average of the total number of active blocks in a plurality of past image frames active_BK: Number of active blocks of the current image frame up to the present

As described above, according to the fifteenth to twentieth aspects, each time one block is encoded, the maximum allowable information amount per block is updated to determine whether the generated information amount is excessive or insufficient. This can be reflected in the maximum allowable information amount of the block, and the generated information amount of the image frame can be made closer to the target information amount or the target buffer occupation amount.

According to the twenty-first and twenty-second inventions, which are dependent on the first and second inventions, the block unit quantization control unit determines the amount of generated information or buffer occupancy per block in advance. If the value is larger than the second threshold value, the encoding of the next block is controlled with a quantization characteristic larger than the set quantization characteristic, and the amount of generated information or the buffer occupancy per block is determined in advance. When the threshold value is smaller than the threshold value of 3, the encoding of the next block is controlled with a quantization characteristic smaller than the set quantization characteristic.

As described above, according to the twenty-first and twenty-second aspects, the second threshold value and the third threshold value are provided.
The quantization characteristic used for encoding the next block is controlled depending on whether the amount of information generated as a result of encoding the block is larger or smaller than each threshold. Thereby, it is possible to cope with a partial change in the movement of the subject or a sudden change in the pattern.

A twenty-third invention is an invention according to the third and fourth inventions, wherein each time one image frame is encoded, a motion amount of a subject included in the image is calculated, and the motion amount is calculated. A motion start / end detection unit that detects one or both of the motion start and the motion end of the subject based on If the total number of blocks is corrected upward and the motion end state is detected,
A motion start / end correction unit that corrects the total number of blocks in the previous image frame downward; As mentioned above, the second
According to the third aspect, by correcting the total number of blocks in the immediately preceding image frame based on a change in the amount of motion of the subject, it is possible to perform encoding control in consideration of the motion of the subject.

A twenty-fourth invention is an invention according to the fifth and sixth inventions, wherein each time one image frame is encoded, a motion amount of a subject included in the image is calculated, and the motion amount is calculated. A motion start / end detection unit that detects one or both of the motion start and the motion end of the subject based on And a motion start / end correction unit that corrects the total number of valid blocks of the previous image frame upward and corrects the total number of valid blocks of the previous image frame downward when the motion end state is detected. As described above, according to the twenty-fourth aspect, by correcting the total number of effective blocks in the immediately preceding image frame based on a change in the amount of motion of a subject, it is possible to perform encoding control in consideration of the motion of the subject. .

According to a twenty-fifth aspect, in accordance with the tenth and eleventh aspects, every time one image frame is encoded, a motion amount of a subject included in the image is calculated, and the motion amount is calculated. A movement start / end detection unit for detecting one or both of the movement start and the movement end of the subject, and a movement start / end detection unit,
Motion start / end correction for correcting the average of the total number of valid blocks of the previous image frame upward when the start of motion state is detected and correcting the average of the total number of valid blocks of the previous image frame downward when the end of motion state is detected And a unit. As described above, according to the twenty-fifth aspect, by correcting the average total number of effective blocks based on the change in the amount of motion of the subject, it is possible to perform encoding control in consideration of the motion of the subject.

A twenty-sixth invention is an invention according to the fifth and sixth inventions, wherein, when the total number of valid blocks of the current image frame exceeds the total number of valid blocks of the previous image frame, a predetermined value is set. The correction is performed by adding to the total number of effective blocks of the previous image frame. As described above, according to the twenty-sixth aspect, by correcting the total number of effective blocks by a predetermined value, even if the motion of the subject in the current image frame is larger than the motion of the subject in the previous image frame, the current image Encoding control can be performed to bring the amount of generated information of the frame closer to the target generated information amount or the target buffer occupation amount. Further, it is possible to perform encoding control without any trouble (runaway software).

A twenty-seventh invention is an invention according to the fifth and sixth inventions, wherein the invalid block of the current image frame (a block in which a coding coefficient is not generated when the image frame is coded). When the total number exceeds the total number of invalid blocks in the previous image frame, correction is performed by subtracting a predetermined value from the total number of valid blocks in the previous image frame. As described above, according to the twenty-seventh aspect, by correcting the total number of valid blocks by a predetermined value,
Even when the motion of the subject in the current image frame is smaller than the motion of the subject in the previous image frame, it is possible to perform the encoding control to make the generated information amount of the image frame closer to the target generated information amount or the target buffer occupation amount.

A twenty-eighth invention is an invention according to the tenth and eleventh inventions, wherein when the total number of valid blocks in the current image frame exceeds the average of the total number of valid blocks in a plurality of past image frames, The correction is performed by adding a determined value to the average of the total number of effective blocks in a plurality of past image frames. As mentioned above, the second
According to the eighth aspect, by correcting the average total number of effective blocks by a predetermined value, even when the motion of the subject in the current image frame is larger than the motion of the subject in a plurality of past image frames, the current image frame is corrected. It is possible to perform encoding control for bringing the generated information amount closer to the target generated information amount or the target buffer occupation amount. Further, it is possible to perform encoding control without any trouble (runaway software).

A twenty-ninth invention is a invention according to the tenth and eleventh inventions, wherein the total number of invalid blocks in the current image frame exceeds the average of the total number of invalid blocks in a plurality of past image frames. The correction is performed by subtracting a predetermined value from the average of the total number of effective blocks in a plurality of past image frames. As described above,
According to the twenty-ninth aspect, by correcting the total number of effective blocks by a predetermined value, even when the motion of the subject in the current image frame is smaller than the motion of the subject in a plurality of past image frames, the generation information of the image frame is obtained. Coding control can be performed to bring the amount closer to the target generated information amount or the target buffer occupation amount.

A thirtieth invention is a invention according to the twenty-fourth to twenty-fourth inventions, wherein the amount of motion of the subject is calculated as a result of a motion vector compensation process used when interframe predictive coding of an image frame is performed. It is characterized in that it is the sum of the absolute values of the output motion vectors. As described above, according to the thirtieth aspect, the motion amount of the subject can be calculated without requiring any special processing.

A thirty-first aspect is an invention according to the first to thirty aspects, wherein the blocks are macroblocks. As described above, according to the thirty-first aspect, the blocks are defined in ITU-T Recommendation H.264. By using a macroblock as a minimum coding unit in the interframe prediction coding method such as H.261, coding control can be performed without special processing such as block separation.

A thirty-second invention is an encoding control method for controlling encoding of an image frame divided into a plurality of blocks by setting a quantization characteristic for each block. The quantization information serving as a reference used for encoding the (current image frame) is represented by generation information indicating the amount of data generated by encoding the immediately preceding image frame (previous image frame) with respect to the current image frame. Deciding based on the quantity, and each time one of the blocks making up the current image frame is encoded,
Calculating the amount of generated information per block, comparing the predetermined threshold value with the calculated amount of generated information per block, based on a result of the comparison and a reference quantization characteristic. Setting a quantization characteristic used for encoding a block next to the encoded block.

A thirty-third invention is a coding control method for controlling the coding of an image frame divided into a plurality of blocks by setting a quantization characteristic for each block. The reference quantization characteristic used for encoding the (current image frame) is used for data transmission after encoding the immediately preceding image frame (previous image frame) with respect to the current image frame. Determining based on the data occupancy (buffer occupancy) in the smoothing buffer, and calculating the buffer occupancy per block each time one of the blocks constituting the current image frame is encoded; A step of comparing a predetermined threshold value with the calculated buffer occupancy per block, based on a result of the comparison and a reference quantization characteristic; There are, and a step of setting a quantization characteristic used to encode the next block of encoded blocks.

The thirty-fourth and thirty-fifth inventions are dependent on the thirty-second and thirty-third inventions, respectively, wherein the predetermined threshold value is the maximum allowable information amount per block, and It is characterized by being calculated from the target generated information amount or the target buffer occupation amount and the total number of blocks constituting one image frame.

[0049] The thirty-sixth and thirty-seventh inventions are dependent on the thirty-second and thirty-third inventions, respectively, wherein the predetermined threshold value is a maximum allowable information amount per block. It is characterized by being calculated from the target generated information amount or the target buffer occupancy amount and the total number of valid blocks in the previous image frame.

A thirty-eighth invention is an invention according to the thirty-fourth and thirty-sixth inventions, wherein the target generated information amount is corrected according to the effective transmission rate (throughput).

A thirty-ninth aspect is an invention according to the thirty-eighth aspect, wherein the correction of the target generated information amount is performed in accordance with a ratio of (effective transmission rate / standard transmission rate).

A fortieth invention is an invention according to the thirty-fourth and thirty-sixth inventions, wherein the target amount of generated information is corrected in accordance with the state of the start or end of the movement of the subject included in the image. Features.

The forty-first and forty-second inventions are inventions according to the thirty-sixth and thirty-seventh inventions, respectively, wherein the total number of valid blocks is an average of the total number of valid blocks in a plurality of past image frames. I do.

A forty-third invention is an invention according to the thirty-second and thirty-third inventions, wherein the step of setting the quantization characteristic comprises the step of: It is characterized in that the characteristic is determined to be equal to, larger than, or smaller than the reference quantization characteristic.

A forty-fourth invention is an invention according to the thirty-second and thirty-third inventions, wherein the step of setting the quantization characteristic comprises the step of setting the quantization characteristic of the first encoded block in the image frame. Is used as a reference quantization characteristic, and based on the result of the comparison, the quantization characteristic of a block to be coded next or later is equal to, larger than, or smaller than the previous block. Is determined.

A forty-fifth invention is an invention according to the thirty-second and thirty-third inventions, wherein the step of setting the quantization characteristic is performed within a predetermined setting width centered on the reference quantization characteristic. , The quantization characteristic of the next block to be encoded is determined.

The forty-sixth and forty-seventh inventions are dependent on the thirty-fourth and thirty-fifth inventions, respectively, wherein the maximum allowable information amount is updated every time a block is coded according to the following equation. Features. Max_info = (target_info-total_info) / (total_BK-
code_BK) Max_info: Maximum allowable information amount target_info: Target generated information amount or target buffer occupancy total_info: Generated information amount or buffer occupation amount up to the present time in the current image frame totak_BK: Total number of blocks constituting one image frame codec_BK: Current image frame Number of blocks coded to date

The forty-eighth and forty-ninth inventions are dependent on the thirty-sixth and thirty-seventh inventions, respectively, wherein the maximum allowable information amount is updated every time a block is coded according to the following equation. Features. Max_info = (target_info-total_info) / (pre_active
_BK-active_BK) Max_info: Maximum allowable information amount target_info: Target generated information amount or target buffer occupation amount total_info: Generated information amount or buffer occupation amount in the current image frame up to the present time pre_active_BK: Total number of active blocks in the previous image frame active_BK: Current image Number of valid blocks so far in the frame

The fiftieth and fifty-first inventions are dependent on the forty-first and forty-second inventions, respectively, wherein the maximum allowable information amount is updated each time a block is encoded according to the following equation. Features. Max_info = (target_info-total_info) / (ave_active
_BK-active_BK) Max_info: Maximum allowable information amount target_info: Target generated information amount or target buffer occupation amount total_info: Generated information amount or buffer occupation amount up to the present time in the current image frame ave_active_BK: Average of the total number of active blocks in a plurality of past image frames active_BK: Number of active blocks of the current image frame up to the present

The 52nd and 53rd inventions are dependent on the 32nd and 33rd inventions, respectively, wherein the step of setting the quantization characteristic is such that the amount of information generated per block or the amount of buffer occupancy is determined in advance. If it is larger than the second threshold value, the encoding of the next block is controlled with a quantization characteristic larger than the set quantization characteristic, and the amount of generated information or buffer occupancy per block is determined in advance. If the threshold value is smaller than the third threshold value, encoding of the next block is controlled with a quantization characteristic smaller than the set quantization characteristic.

A fifty-fourth invention is an invention according to the thirty-fourth and thirty-fifth inventions, wherein each time one image frame is encoded, the motion amount of a subject included in the image is calculated, and the motion amount is calculated. Based on the step of detecting one or both of the state of the start or end of movement of the subject, and in the step of detecting, when the start of movement is detected, the total number of blocks of the previous image frame is corrected upward, Correcting the total number of blocks in the previous image frame downward when the motion end state is detected.

A fifty-fifth invention is an invention according to the thirty-sixth and thirty-seventh inventions, wherein each time one image frame is encoded, a motion amount of a subject included in the image is calculated, and the motion amount is calculated. Detecting the state of either or both of the movement start and the movement end of the subject based on the above, and in the step of detecting, when the movement start state is detected, the total number of effective blocks of the previous image frame is corrected upward. And correcting the total number of valid blocks in the previous image frame downward when the motion end state is detected.

A fifty-sixth invention is an invention according to the forty-first and forty-second inventions, wherein each time one image frame is encoded, the motion amount of a subject included in the image is calculated, and the motion amount is calculated. Detecting the state of one or both of the start and end of the movement of the subject based on the above, and in the step of detecting, when the start of movement is detected, the average of the total number of effective blocks of the previous image frame is increased. And correcting the average of the total number of valid blocks in the previous image frame downward when the motion end state is detected.

A fifty-seventh invention is a invention according to the thirty-sixth and thirty-seventh inventions, wherein when the total number of valid blocks of the current image frame exceeds the total number of valid blocks of the previous image frame, a predetermined value is set. The correction is performed by adding to the total number of effective blocks of the previous image frame.

A fifty-eighth invention is an invention according to the thirty-sixth and thirty-seventh inventions, wherein when the total number of invalid blocks in the current image frame exceeds the total number of invalid blocks in the previous image frame, a predetermined value is set. Is subtracted from the total number of effective blocks in the previous image frame to make a correction.

A fifty-seventh invention is an invention according to the forty-first and forty-second inventions, wherein if the total number of valid blocks in the current image frame exceeds the average of the total number of valid blocks in a plurality of past image frames, The correction is performed by adding the determined value to the average of the total number of effective blocks in a plurality of past image frames.

A sixtieth invention is according to the forty-first and forty-second inventions, wherein the total number of invalid blocks in the current image frame exceeds the average of the total number of invalid blocks in a plurality of past image frames. The correction is performed by subtracting a predetermined value from the average of the total number of effective blocks in a plurality of past image frames.

A sixty-first invention is an invention according to the fifty-fourth to fifty-sixth inventions, wherein the amount of motion of the subject is calculated as a result of a motion vector compensation process used when interframe predictive coding of an image frame is performed. It is characterized in that it is the sum of the absolute values of the output motion vectors.

A 62nd invention is an invention according to the 32nd to 61st inventions, wherein the block is a macroblock.

As described above, the 32nd to 62nd inventions
The present invention relates to an encoding control method performed by the encoding control device according to the first to thirty-first aspects. Therefore, the 32nd to 62nd inventions have the same effects as the above 1st to 31st inventions.

According to a sixty-third aspect, an encoding control method for controlling encoding of an image frame divided into a plurality of blocks by setting a quantization characteristic for each block is provided as a program executable by a computer device. On a recorded medium, a quantization characteristic serving as a reference used for encoding an image frame (current image frame) to be processed is defined as:
A step of deciding based on an amount of generated information which is a data amount generated by encoding a previous image frame (previous image frame) with respect to the current image frame; Calculating the amount of generated information per block each time one is encoded; comparing the predetermined threshold with the calculated amount of generated information per block; Setting a quantization characteristic to be used for encoding a block next to the encoded block based on the result and a reference quantization characteristic.

According to a sixty-fourth aspect, an encoding control method for controlling the encoding of an image frame divided into a plurality of blocks by setting a quantization characteristic for each block is provided as a program executable by a computer device. On a recorded medium, a quantization characteristic serving as a reference used for encoding an image frame (current image frame) to be processed is defined as:
Deciding on the basis of the data occupancy (buffer occupancy) in the smoothing buffer used for data transmission of the immediately preceding image frame (previous image frame) after encoding of the current image frame; Calculating a buffer occupancy per block each time one of the blocks constituting the current image frame is coded, a predetermined threshold value, and a calculated buffer occupancy per block. To perform at least a step of comparing, and a step of setting a quantization characteristic used for encoding a block next to the encoded block based on a result of the comparison and a reference quantization characteristic. The program has been recorded.

[0073]

BEST MODE FOR CARRYING OUT THE INVENTION Hereinafter, ITU-T Recommendation H. 261
The case where the coding control method according to each embodiment of the present invention is applied to the CIF image of FIG. As we all know,
The CIF image is composed of 352 × 288 dots.
In H.261, this is divided into 22 × 18 blocks in units of 16 × 16 dot macroblocks.
In the present embodiment, the blocks used for the description are ITU-T
Recommendation H. It is assumed that the macroblock is a minimum coding unit in the interframe prediction coding of H.261. The block that can be used in the present invention is not limited to this macro block, and a block having another dot size may be used.

(First Embodiment) FIG. 1 shows a first embodiment of the present invention.
1 is a block diagram illustrating a configuration of an image encoding device 1 including an encoding control device 11 that uses an encoding control method according to an embodiment. In FIG. 1, the image encoding device 1 includes an encoding control device 11 and an image encoding unit 12. The image encoding unit 12 compresses and encodes the video signal Sv input to the image encoding device 1 to generate encoded data Dc based on the control of the encoding control device 11, and generates the encoded data Dc.
c is output from the image encoding device 1. Coding control device 1
Reference numeral 1 denotes a frame unit quantization control unit 111, a block unit generated information amount calculation unit 112, a comparison unit 113, a maximum allowable information amount calculation unit 114, and a block unit quantization control unit 1.
The encoding control method according to the first embodiment of the present invention controls the encoding of the video signal Sv performed by the image encoding unit 12. First, the operation of the encoding control device 11 according to the first embodiment will be described in detail.

The frame unit quantization control unit 111 inputs the amount of generated information, which is the amount of data generated by encoding one image frame, from the image encoding unit 12. Then, the frame unit quantization control unit 111 sets the quantization characteristic used for encoding the next image frame to be processed (hereinafter, referred to as the current image frame) to the image frame immediately before the current image frame (the current image frame). Hereinafter, referred to as the previous image frame)
Is determined based on the amount of generated information, and is output to the block-unit quantization control unit 115. The quantization characteristic used for encoding the current image frame is determined by the occupation amount (buffer occupation amount) of image data in a smoothing buffer (not shown) used for data transmission after encoding. You may. The method of determining the quantization characteristic based on the buffer occupancy can take into account the effective transmission rate (throughput) that fluctuates due to retransmission or the like, and is particularly useful for an image encoding device having a retransmission control function. Further, the determination of the quantization characteristic in the frame unit quantization control unit 111 is not limited to the method described in the present embodiment, but may be performed by a conventional coding control method (for example, Japanese Patent Laid-Open No. 7-1 described in the related art).
No. 07482).

The block unit generated information amount calculation unit 112
The amount of generated information up to the current point in the current image frame being coded, that is, the amount of generated information obtained by accumulating the amount of generated information of the coded block among a plurality of blocks constituting the current image frame To the image encoding unit 12
Enter from. Then, the block unit generated information amount calculation unit 112 calculates the generated information amount per block from the input cumulative generated information amount and outputs the calculated information amount to the comparison unit 113.

Each time one block in the current image frame is coded by the image coding unit 12, the maximum allowable information amount calculation unit 114 calculates the remaining one block per block to be coded according to the following equation (1). Max_in
Calculate fo. Therefore, the maximum allowable information amount Max_info is 1
The value will be sequentially updated to a value calculated each time a block is encoded. Max_info = (target_info−total_info) / (total_BK−codec_BK) (1) target_info: predetermined target generated information amount total_info: generated information amount up to the present time in the current image frame total_BK: total number of blocks constituting the image frame codec_BK: Number of blocks of the current image frame that have been coded up to the present time. In the case where the frame unit quantization control unit 111 determines the quantization characteristic based on the buffer occupancy, the numerator on the right side of the above equation (1) is (Predetermined target buffer occupancy)-(buffer occupancy up to the present time). This calculated maximum allowable information amount Ma
x_info is output to comparing section 113.

Each time the image encoding unit 12 encodes one block in the current image frame, the comparing unit 113
The generated information amount per block calculated by the block generated information amount calculation unit 112 is compared with the maximum allowable information amount calculated by the maximum allowable information amount calculation unit 114, and the comparison result (coding control error ) Is output to the block unit quantization control unit 115.

The block unit quantization control unit 115 determines a quantization characteristic used for coding the next block based on the coding control error output from the comparison unit 113,
Output to the image encoding unit 12. At this time, the block unit quantization control unit 115
The quantization characteristic used for encoding the next block is variably determined within a predetermined range centered on the quantization characteristic determined in (1). Specifically, the block unit quantization control unit 115 sets the quantization characteristics determined by the frame unit quantization control unit 111 as quantization characteristics used for encoding the first block in the current image frame.
Then, for the next and subsequent blocks, if the comparison result indicates that the amount of generated information per block is larger than the maximum allowable information amount, the block unit quantization control unit 115 determines the quantization characteristic used for encoding the next block. Is set to a quantization characteristic (improving image quality) smaller than the quantization characteristic used for encoding the previous block. In addition, when the comparison result indicates that the amount of information generated per block is smaller than the maximum allowable information amount, the block unit quantization control unit 115 determines the quantization characteristic used for encoding the next block as:
The quantization characteristics are set to be larger (the image quality is deteriorated) than the quantization characteristics used for encoding the previous block. In addition,
How small or large the quantization characteristic is set can be arbitrarily determined according to the performance given to the image encoding device 1.

Next, an encoding control method performed by the encoding control device 11 according to the first embodiment of the present invention will be described with reference to FIG. FIG. 2 is a flowchart illustrating processing performed by the encoding control method according to the first embodiment of the present invention. First, the total number of blocks of one image frame and the target amount of generated information are set (step S201).
In the present embodiment, the image format is defined in ITU-T Recommendation H.264. 261 CIF images, so the total number of blocks constituting one image frame (the total number of macroblocks)
Stores the value 396. The target amount of generated information is 1
The target generated information amount in the image frame is stored.
For example, when transmitting a moving image of 30 frames / second at a transmission rate of 1 Mbps, 33333 bits are stored. Next, a current image frame to be encoded is captured (step S202). This target generated information amount may be fixed or may be dynamically corrected according to the effective transmission speed (throughput) or the like. For example, the corrected target generated information amount is calculated according to the corrected target generated information amount = target generated information amount × (effective transmission speed / standard transmission speed). With such correction, even when a transmission error occurs due to deterioration of the radio wave condition or the like and the effective transmission speed changes due to retransmission or the like, optimal encoding control according to the radio wave condition can be performed. Next, it is determined whether all the blocks in one image frame, that is, all the 396 blocks have been encoded (step S).
203).

If it is determined in step S203 that all blocks have been coded, the current image frame is coded based on the target generated information amount and the generated information amount obtained when the previous image frame is coded. Is set (step S204).
In addition, variables storing the generated information amount and the maximum allowable information amount per block are initialized (cleared to zero), respectively (step S205). On the other hand, step S2
If it is determined in step 03 that all blocks have not been coded, the remaining next block is coded using the quantization characteristic Quant obtained in the processing of the previous block (step S206). Next, the amount of generated information per block is calculated (step S207).
Specifically, the difference between the accumulated value of the generated information amount of the block coded up to the present time and the accumulated value of the generated information amount of each block coded up to one block before is obtained, Is generated per block. Next, the maximum allowable information is obtained from the difference between the target generated information amount and the generated information amount in the current image frame up to the present time, and the difference between the total number of blocks constituting the current image frame and the number of blocks coded so far. The amount is calculated (Step S208).

Next, the generated information amount per block is compared with the maximum allowable information amount (step S209). If the generated information amount is larger in step S209,
Quantization characteristic Quan used for encoding the previous block
t is incremented, and the quantization characteristic used for encoding the next block is set to be large (in a direction in which the image quality deteriorates) (step S210). On the other hand, step S20
9, if the amount of generated information is smaller, the quantization characteristic Quant used for encoding the previous block is decremented, and the quantization characteristic used for encoding the next block is reduced (in the direction in which the image quality is improved). Is set (step S211).

As described above, according to the coding control apparatus and method according to the first embodiment of the present invention, the quantization characteristic used for coding each block is determined by the frame unit quantization control unit 111. Is set in the vicinity of the obtained quantization characteristic.
As a result, an image frame can be encoded using a quantization characteristic whose quantization value can be changed in a minimum unit (that is, “1”) or less. You can get closer. Therefore, it is possible to improve the accuracy of the encoding control, reduce the margin of the target generated information amount as much as possible, and perform the encoded image transmission using the transmission band efficiently.

The block unit quantization control unit 115
Has a width (upper limit value and lower limit value) for determining whether or not to change the quantization characteristic, so that an image with little change in image quality (without image quality unevenness) can be obtained for each block.
Further, the maximum allowable information amount calculation unit 114 calculates (updates) the maximum allowable information amount per block every time one block is coded, so that the amount of generated information due to an error in coding control is excessive or insufficient. Can be reflected in the maximum allowable information amount for determining the quantization characteristic used for encoding the next block, and the generated information amount of the image frame can be made closer to the target information amount. Further, a block, which is a unit for performing coding control, is described in ITU-T Recommendation H. By using a macroblock as the minimum coding unit in the inter-frame predictive coding method such as H.261, special processing such as processing for separating blocks is not required, and coding control can be easily performed.

In the first embodiment, the block unit quantization control unit 115 determines the code of the next block within a predetermined range centering on the quantization characteristic determined by the frame unit quantization control unit 111. Although it has been described that the quantization characteristic used for encoding is determined, the quantization characteristic used for encoding the next block may be determined without having a predetermined range. In this case, the same effect as described above can be obtained. In the first embodiment, based on the comparison result output from the comparison unit 113, the block-unit quantization control unit 115 changes the quantization characteristic used for encoding the next block into the code of the previous block. It has been described that the quantization characteristics are always set to be larger or smaller than the quantization characteristics used for the quantization. However, the block unit quantization control unit 115 is provided with a predetermined coding control error width that does not change the quantization characteristic, and the difference between the generated information amount per block and the maximum allowable information amount (the coding control error ) Is within this range, the quantization characteristic may not be changed. In this case, since the change in the quantization characteristic for each block is reduced, stable image quality can be obtained (the image quality does not differ for each block), and the amount of transmission of the change information of the quantization characteristic is reduced. , The compression ratio can be improved.

Further, the block unit quantization control unit 115 in the first embodiment determines the code of the next block in a fixed change width based on the minimum unit (ie, “1”) based on the comparison result of the comparison unit 113. Quantization characteristics used for encoding are set, but a threshold value is set in advance, and the difference between the amount of generated information per block and the maximum allowable information amount (encoding control error) exceeds the threshold value. The change width of the quantization value may be changed depending on whether or not. For example, when the encoding control error exceeds a predetermined threshold, the quantization characteristic used for encoding the next block is set to the quantization characteristic used for encoding the previous block +2, It is assumed that the quantization characteristic used for coding the block is −2. By doing so, it is possible to perform encoding control that more closely follows the movement of the subject. Further, in the first embodiment, each time one block is encoded, the maximum allowable information amount calculation unit 114 calculates the difference between the amount of information generated as a result of actually encoding the block and the predicted information amount. Although the maximum allowable information amount is calculated (updated) in order to correct the error, the calculation is not necessarily required. In this case, the processing time required for calculation and the software size can be reduced.

(Second Embodiment) FIG. 3 shows a second embodiment of the present invention.
It is a block diagram which shows the structure of the image encoding device 3 including the encoding control device 31 which uses the encoding control method which concerns on embodiment. 3, the image encoding device 3 includes an encoding control device 31 and an image encoding unit 12. The image encoding unit 12 has the same configuration as that of the first embodiment, and a description thereof will be omitted. The encoding control device 31 includes a frame unit quantization control unit 111, a block unit generated information amount calculation unit 112, a comparison unit 113, a maximum allowable information amount calculation unit 311, a movement start / end detection unit 312, The image encoding unit 12 includes a start / end correction unit 313 and a block-based quantization control unit 115, and performs encoding of the video signal Sv performed by the image encoding unit 12 using the encoding control method according to the second embodiment of the present invention. Control. Note that the frame unit quantization control unit 1 of the encoding control device 31 according to the second embodiment
11. Block unit generated information amount calculation unit 112, comparison unit 1
13 and the block-unit quantization control unit 115 have the same configuration as the coding control device 11 according to the first embodiment, and thus description thereof will be omitted. First, the operation of the encoding control device 31 according to the second embodiment will be described in detail.

The motion start / end detection unit 312 inputs a motion vector obtained by the motion compensation processing from the image encoding unit 12 every time one image frame is encoded. Then, the movement start / end detection unit 312 detects the movement start state and / or the movement end state of the subject based on the motion vector. Specifically, the motion start / end detection unit 312 calculates the sum of the absolute values of the motion vectors of each block every time one image frame is encoded. Then, the movement start / end detection unit 312
Determines that the subject is in the starting state if the sum of the absolute values of the motion vectors over a plurality of past image frames is increasing, and determines that the subject is in the ending state if the total is lower. This determination result is output to the movement start / end correction unit 313.

The movement start / end correction unit 313 corrects the total number of blocks of the previous image frame according to the movement of the subject detected by the movement start / end detection unit 312. Specifically, when it is determined that the subject is in the state of starting to move, the movement start / end correction unit 313 determines that there is a high possibility that the amount of generated information increases after the current image frame, and Correction is made in the direction to increase the total number of blocks. In addition, when it is determined that the subject is in the motion end state, the movement start / end correction unit 313 determines that there is a high possibility that the amount of generated information will decrease after the current image frame, and determines the total number of blocks in the previous image frame. Correct in the direction to decrease. Note that the extent to which the total number of blocks is corrected can be arbitrarily determined according to the performance given to the image encoding device 1. The correction value of the total number of blocks may be a predetermined fixed value or a value that dynamically changes according to the amount of movement of the subject. In the latter case, it is possible to perform coding control in consideration of the movement of the subject. In the second embodiment, a case where a fixed correction value is used will be described.

Each time one block in the current image frame is encoded by the image encoding unit 12, the maximum allowable information amount calculation unit 311 calculates the maximum allowable information amount in accordance with the above-described equation (1), similarly to the maximum allowable information amount calculation unit 114. , The maximum allowable information amount Max_info per block remaining to be encoded is calculated. When calculating the maximum allowable information amount Max_info,
The total number of blocks total_BK constituting the current image frame corrected by the movement start / end correction unit 313 is applied to the above equation (1). This calculated maximum allowable information amount
Max_info is output to comparing section 113.

Next, an encoding control method performed by the encoding control device 31 according to the second embodiment of the present invention will be described. FIG. 4 is a flowchart illustrating processing performed by the encoding control method according to the second embodiment of the present invention. Note that, in FIG. 4, the processes (steps S202 to S211) having the same step numbers as those in FIG. .

First, the total number of blocks of one image frame, the amount of target occurrence information, and the correction value of the total number of blocks are set (step S401). In the present embodiment, the image format is defined in ITU-T Recommendation H.264. Since it is assumed that the CIF image is a CIF image of H.261, a value 396 is stored in the total number of blocks (the total number of macroblocks) constituting one image frame.
The target generated information amount stores a target generated information amount in one image frame. For example, transmission rate 1Mbps
When transmitting a moving image of 30 frames / second by using
33 bits will be stored. This target generated information amount may be fixed, or may be dynamically corrected according to the effective transmission speed (throughput) as described above. Further, the correction may be dynamically made according to the amount of movement of the subject. For example, the target generated information amount is corrected to a small amount when the subject starts moving, and to a large amount when the subject ends moving. With this correction, the amount of generated information of each image frame is controlled to be close to the target amount of generated information, and optimal coding control according to the amount of motion of the subject can be performed. The block total correction value is a value for correcting the total number of blocks constituting the current image frame in accordance with the amount of motion of the subject, and affects the calculation of the maximum allowable information amount.

In step S205, when the block-based generated information amount and the maximum allowable information amount are initialized, the start or end of the movement of the subject or both are detected (step S402). Specifically, in a motion vector search process, which is one process for encoding a block, the absolute value of the motion vector as a detection result is cumulatively added for each block, and the absolute value sum of the motion vectors of the current image frame is calculated. Is calculated. Then, the total sum of the absolute values of the motion vectors is stored for a plurality of past image frames, and if the total sum of the absolute values is increasing, it is determined that the subject starts to move. Judge that the movement is over.

Next, it is determined whether or not the current image frame on which the encoding process has been completed is related to the start of the movement of the subject and whether it is related to the end of the movement of the subject (steps S403 and S404). If it is determined in steps S403 and S404 that the current image frame is related to the start of the movement of the subject, the block total correction value is added to the block total so that the calculated maximum allowable information amount becomes smaller. (Step S4
05). Such correction is performed after the current image frame.
This is based on the prediction that the movement of the subject is further increased and the amount of generated information tends to increase. If it is determined in steps S403 and S404 that the current image frame is related to the end of the movement of the subject, the correction value for the total number of blocks is subtracted from the total number of blocks so that the calculated maximum allowable information amount increases. (Step S406). Such correction is based on the prediction that after the current image frame, the motion of the subject becomes smaller and the amount of generated information tends to decrease.
On the other hand, if it is determined in steps S403 and S404 that the current image frame does not relate to the start / end of the movement of the subject, the process returns to step S202 without correcting the total number of blocks, and the encoding process of the next image frame is performed. Move on to

As described above, according to the encoding control device and method according to the second embodiment of the present invention, the first
Movement start / end correction unit 3 in addition to the effects of the first embodiment.
By correcting the total number of blocks of the previous image frame by the method 13, encoding control can be performed in consideration of the motion of the subject. In addition, since the motion amount of the subject is calculated using the motion vector obtained from the motion vector compensation process performed when the image frame is subjected to the inter-frame predictive encoding, a special process only for calculating the motion amount of the subject is performed. Do not need.

In the second embodiment, each time one block is encoded, the maximum permissible information amount calculation unit 311 calculates the generated information amount as a result of actually encoding the block and the predicted information amount. Although the maximum allowable information amount is calculated (updated) in order to correct an error with the amount, it is not always necessary to calculate the maximum allowable information amount. In this case, the processing time required for calculation and the software size can be reduced.

(Third Embodiment) FIG. 5 shows a third embodiment of the present invention.
It is a block diagram which shows the structure of the image encoding apparatus 5 including the encoding control apparatus 51 which uses the encoding control method which concerns on embodiment. In FIG. 5, the image encoding device 5 includes an encoding control device 51 and an image encoding unit 12. Since the image encoding unit 12 has the same configuration as the first and second embodiments, description thereof will be omitted. Encoding control device 51
The frame unit quantization control unit 111, the block unit generated information amount calculation unit 112, the comparison unit 113, the valid block calculation unit 511, the invalid block calculation unit 512, the maximum allowable information amount calculation unit 513, Block correction unit 51
4, a movement start / end detection unit 312, a movement start / end correction unit 515, and a block unit quantization control unit 115.
And the video signal S performed by the image encoding unit 12 by the encoding control method according to the third embodiment of the present invention.
controls the encoding of v. Note that the frame unit quantization control unit 111 of the coding control device 51 according to the third embodiment
The block unit generated information amount calculation unit 112, the comparison unit 113, and the block unit quantization control unit 115 have the same configuration as the encoding control device 11 according to the first embodiment, and the movement start / end detection unit 312 Since the configuration is the same as that of the encoding control device 31 according to the second embodiment, the description is omitted here. First, the operation of the encoding control device 51 according to the third embodiment will be described in detail.

The effective block calculating section 511 inputs the presence / absence of a coding coefficient for each block to be coded from the image coding section 12, and calculates the total number of valid blocks per image frame. The effective block means a block in which a coding coefficient has been generated when the image is coded by the image coding unit 12. Effective block calculator 51
Specifically, every time a block is coded, it is determined whether or not there is a coding coefficient for the block, and the number of blocks having the coding coefficient is counted.
Calculate the total number of valid blocks.

The invalid block calculating unit 512 receives the presence or absence of a coding coefficient for each block to be coded from the image coding unit 12, and calculates the total number of invalid blocks per one image frame. The invalid block means a block in which a coding coefficient has not been generated when the coding is performed by the image coding unit 12. Specifically, every time a block is coded, the invalid block calculation unit 512 determines whether or not there is a coding coefficient for the block, and counts the number of blocks having no coding coefficient. , Calculate the total number of invalid blocks.

Each time one block in the current image frame is encoded by the image encoding unit 12, the maximum allowable information amount calculation unit 513 calculates, for each of the remaining one block to be encoded according to the following equation (2). Max_in information amount Max_in
Calculate fo. Therefore, the maximum allowable information amount Max_info is 1
The value will be sequentially updated to a value calculated each time a block is encoded. Max_info = (target_info−total_info) / (pre_active_BK−active_BK) (2) target_info: predetermined target generated information amount total_info: generated information amount in the current image frame up to the present time pre_active_BK: total number of active blocks in the previous image frame active_BK : The number of effective blocks of the current image frame up to the present time. In the case where the frame unit quantization control unit 111 determines the quantization characteristic based on the buffer occupancy, the right side numerator of the above equation (2) is expressed by the following equation. (Target buffer occupancy)-(buffer occupancy up to the present time). When calculating the maximum allowable information amount Max_info, the total number of effective blocks pre_active_BK of the previous image frame corrected by the movement start / end correction unit 515 or the effective block correction unit 514 is applied to the above equation (2). This calculated maximum allowable information amount Ma
x_info is output to comparing section 113.

The movement start / end correction unit 515 corrects the total number of effective blocks in the previous image frame according to the movement of the subject detected by the movement start / end detection unit 312.
Specifically, the movement start / end correction unit 515 determines that there is a high possibility that the number of effective blocks will increase after the current image frame when it is determined that the subject is in the movement start state. Correction is performed in a direction to increase the total number of valid blocks. In addition, the movement start / end correction unit 5
When it is determined that the subject is in the motion end state, the determination unit 15 determines that there is a high possibility that the number of valid blocks will decrease after the current image frame, and corrects the direction to reduce the total number of valid blocks of the previous image frame. Note that how much the total number of effective blocks is corrected can be arbitrarily determined according to the performance given to the image encoding device 1. The correction value of the total number of effective blocks may be a fixed value that is determined in advance or a value that dynamically changes according to the amount of movement of the subject. In the latter case, it is possible to perform coding control in consideration of the movement of the subject. In the third embodiment, a case where a fixed correction value is used will be described.

The effective block correction unit 514 determines that the number of effective blocks in the current image frame is
If the total number of valid blocks in the previous image frame calculated in step 11 is larger than the total number of valid blocks in the previous image frame, a predetermined value (1 in this example) is corrected. Also, an effective block correction unit 514
When the number of invalid blocks in the current image frame is larger than the total number of invalid blocks in the previous image frame calculated by the invalid block calculating unit 512, a predetermined value (1 in this example) is set to the previous image frame. Is corrected by subtracting from the total number of effective blocks.

Next, an encoding control method performed by the encoding control device 51 according to the third embodiment of the present invention will be described. FIG. 6 is a flowchart illustrating processing performed by the encoding control method according to the third embodiment of the present invention. In FIG. 6, the processes denoted by the same step numbers as those in FIGS. 2 and 4 (steps S202 to S20).
4, S206, S207, S209 to S211, S40
Steps 2 to S404) are the same as the processing described in the first and second embodiments, and a description thereof will be omitted.

First, the setting of the target generated information amount, the setting of the effective block total correction value, and the initialization of the variables storing the total number of effective blocks and the total number of invalid blocks (clearing to zero) are as follows:
Each is performed (step S601). The target generated information amount stores a target generated information amount in one image frame. For example, when transmitting a moving image of 30 frames / second at a transmission rate of 1 Mbps, 33333 bits are stored. The effective block total correction value is a value for correcting the total number of effective blocks of the previous image frame when there is a difference between the motion of the subject in the previous image frame and the motion of the subject in the current image frame. Affects the amount of information. The effective block total correction value is also a value for correcting the total number of effective blocks of the previous image frame according to the amount of movement of the subject. In the present embodiment, when there is a difference between the motion of the subject in the previous image frame and the motion of the subject in the current image frame, the value of the effective block total number of the previous image frame is corrected and the motion amount of the subject is changed. Although the value for correcting the total number of effective blocks of the previous image frame is the same value, they may be different values. By setting different values, encoding control accuracy can be improved.

When the quantization characteristic Quant is set in step S204, the variables storing the amount of information generated per block and the maximum allowable information amount are each initialized (cleared to zero) (step S602).
Further, the total number of valid blocks of the current image frame that has just been encoded is stored as the total number of previous valid blocks, and the total number of invalid blocks of the current image frame that has just completed encoding is stored as the total number of previous invalid blocks, respectively (step S60).
2). Next, in the above steps S403 and S404, when it is determined that the current image frame is related to the start of movement of the subject, the previous effective block total number is corrected to the previous effective block total number so that the calculated maximum allowable information amount becomes smaller. The value is added (step S603). Such correction is based on the prediction that after the current image frame, the movement of the subject becomes larger and the amount of generated information tends to increase. Step S403,
If it is determined in step S404 that the current image frame is related to the end of the movement of the subject, the correction value of the total number of valid blocks is subtracted from the total number of previous valid blocks so that the calculated maximum allowable information amount becomes large ( Step S604). Such correction is based on the prediction that after the current image frame, the motion of the subject becomes smaller and the amount of generated information tends to decrease.

In step S207, when the amount of generated information in blocks is calculated, the difference between the target amount of generated information and the amount of generated information up to the present time and the total number of valid blocks in the previous image frame (previous valid block) The maximum allowable information amount is calculated from the difference between the total number) and the number of valid blocks of the current image frame up to the present (step S60).
5). Next, the comparison between the total number of previous valid blocks and the current number of valid blocks and the comparison between the total number of previous invalid blocks and the current number of invalid blocks are performed (steps S606 and S607). This step S60
In 6, when it is determined that the total number of previous valid blocks is smaller, the total number of valid blocks is added to the total number of previous valid blocks to correct the total number of previous valid blocks (step S608). That is, the case where step S608 is executed is a case where the motion of the current image frame is larger than the motion of the previous image frame. Step S
If it is determined in 607 that the total number of previous invalid blocks is smaller, the correction value of the total number of valid blocks is subtracted from the total number of previous valid blocks to correct the total number of previous valid blocks (step S609). That is, this step S6
09 is executed when the motion of the current image frame is smaller than the motion of the previous image frame.

As described above, according to the encoding control device and method according to the third embodiment of the present invention, the first
In addition to the effects of the second embodiment, the maximum allowable information amount calculation unit 513 updates the maximum allowable information amount per block every time one block is encoded. This allows
The excess or deficiency of the amount of information generated due to an error in coding control can be reflected in the maximum allowable information amount that determines the quantization characteristic of the next block, and the amount of information generated in an image frame can be made closer to the target amount of generated information. . In addition, the motion start / end correction unit 515 corrects the total number of effective blocks in the previous image frame based on a change in the amount of motion of the subject, thereby performing encoding control in consideration of the motion of the subject.
Further, the valid block correction unit 514 corrects the total number of valid blocks of the previous image frame based on the total number of valid blocks or the total number of invalid blocks of the current image frame. Thereby, even if a difference occurs in the total number of effective blocks between the previous image frame and the current image frame due to a change in the movement of the subject or the like, the amount of generated information of the image frame can be made closer to the target amount of generated information. Further, it is possible to perform coding control without any trouble (runaway software).

In the third embodiment, every time one block is encoded, the maximum permissible information amount calculation unit 513 determines the amount of information generated as a result of actually encoding the block and the predicted information amount. Although the maximum allowable information amount is calculated (updated) in order to correct an error with the amount, the calculation is not necessarily required. In this case, the processing time required for calculation and the software size can be reduced. When calculating the maximum allowable information amount in the maximum allowable information amount calculation unit 513, the total number of effective blocks pre_ac
Instead of tive_BK, the average total number of effective blocks obtained from a plurality of past image frames may be used. By using such an average effective block total number, it is possible to calculate the maximum allowable information amount per block that is not affected by the temporary movement of the subject. Further, in the effective block correction unit 514, the total number of valid blocks of the previous image frame used for determining whether or not to correct the total number of valid blocks is set to:
An average total number of effective blocks obtained from a plurality of past image frames may be used. Similarly, as the total number of invalid blocks, the average total number of invalid blocks obtained from a plurality of past image frames may be used. By using the average total number of effective blocks and the average total number of invalid blocks, it is possible to perform coding control that is not affected by the temporary movement of the subject.

It should be noted that each configuration of the image encoding devices 1, 3 and 5 shown in each of the above embodiments is different from the encoding control devices 11 and 3
1, 51 may be realized by hardware, but may also be realized by using software. In the latter case, for example, CPU (Central Processing Un
it), using a memory and an external storage device, the CPU executes a predetermined program (image encoding program and encoding control program) loaded from the external storage device into the memory, and thereby the image encoding devices 1, 3, 5 can be realized. In this case, these programs are typically provided by a recording medium (flexible disk, CD-ROM, DVD, or the like) on which the programs are recorded. That is, the user sets the purchased recording medium in the image encoding devices 1, 3, and 5 and causes the image encoding devices 1, 3, and 5 to read the program stored therein, thereby storing the program in the hard disk or the like. Install on the device. Alternatively, a program transmitted online to the image encoding devices 1, 3, and 5 via a communication line may be installed in the storage device. Further, the program may be installed in a storage device before the manufacturer ships the image encoding devices 1, 3, and 5. The program thus installed is loaded from the storage device into the memory and executed by the CPU.

[Brief description of the drawings]

FIG. 1 is a block diagram illustrating a configuration of an image encoding device 1 including an encoding control device 11 using an encoding control method according to a first embodiment of the present invention.

FIG. 2 is a flowchart illustrating processing performed by an encoding control method according to the first embodiment of the present invention.

FIG. 3 is a block diagram illustrating a configuration of an image encoding device 3 including an encoding control device 31 using an encoding control method according to a second embodiment of the present invention.

FIG. 4 is a flowchart illustrating processing performed by an encoding control method according to a second embodiment of the present invention.

FIG. 5 is a block diagram showing a configuration of an image encoding device 5 including an encoding control device 51 using an encoding control method according to a third embodiment of the present invention.

FIG. 6 is a flowchart illustrating processing performed by an encoding control method according to a third embodiment of the present invention.

FIG. 7 is a block diagram illustrating an example of a configuration of a conventional encoding control device.

FIG. 8 is a diagram illustrating an example of an operation of a conventional encoding control device.

[Explanation of symbols]

 1, 3, 5 ... image coding device 11, 31, 51 ... coding control device 12 ... image coding unit 111 ... frame unit quantization control unit 112 ... block unit generated information amount calculation unit 113 ... comparison unit 114, 311 , 513: maximum allowable information amount calculation unit 115: block unit quantization control unit 312: movement start / end detection unit 313, 515: movement start / end correction unit 511: valid block calculation unit 512: invalid block calculation unit 514: valid Block corrector 700 Encoder 701 Subtractor 702 Orthogonal transformer 703 Quantizer 704, 751 Inverse quantizer 705, 752 Inverse orthogonal transformer 706, 753 Adder 707, 754 Image memory 708, 755 ... intra-loop filter 709 ... encoding controller 710,711,756 ... selector 750 ... decoder

Continued on the front page F-term (reference) 5C059 KK06 KK22 MA01 MA23 MD02 PP04 RA08 SS07 TA52 TB08 TC18 TC41 TD06 TD11 UA02 5C064 AA01 AA02 AB04 AC02 AD02 AD08 AD14 5J064 AA01 AA02 BA13 BA15 BC02 BC14 BC16 BC27 BD03

Claims (64)

[Claims] An encoding control device controls the encoding of an image frame divided into a plurality of blocks by setting a quantization characteristic for each of the blocks. A reference quantization characteristic used for encoding an image frame) is a data amount generated by encoding an image frame immediately before the current image frame (hereinafter, referred to as a previous image frame). Frame-based quantization control means for determining based on the generated information amount; and each time one of the blocks constituting the current image frame is encoded, generated information amount calculation for calculating the generated information amount per block. Means for comparing a predetermined threshold value with the amount of generated information per block calculated by the generated information amount calculating means; A block-based quantization control unit that sets a quantization characteristic used for encoding a block next to the encoded block based on the comparison result and the reference quantization characteristic. . 2. An encoding control apparatus for controlling encoding of an image frame divided into a plurality of blocks by setting a quantization characteristic for each block, wherein the image frame to be processed from now on (hereinafter referred to as the current frame). A quantization characteristic which is a reference used for encoding an image frame) is used for data transmission after encoding the immediately preceding image frame (hereinafter, referred to as a previous image frame) with respect to the current image frame. Frame-based quantization control means for determining based on the data occupancy in the smoothing buffer (hereinafter referred to as buffer occupancy), and each time one of the blocks constituting the current image frame is encoded, Generating information amount calculating means for calculating a buffer occupancy amount per block; a predetermined threshold value; and one block corresponding to the one block calculated by the generating information amount calculating means. A comparison means for comparing the buffer occupancy of the block with the data, and a quantization used for encoding a block next to the encoded block based on a comparison result of the comparison means and the reference quantization characteristic. An encoding control device comprising: a block-based quantization control unit that sets characteristics. 3. The predetermined threshold value is a maximum allowable information amount per block, and is calculated from a predetermined target generated information amount and a total number of blocks constituting one image frame. The encoding control device according to claim 1, characterized in that: 4. The predetermined threshold value is a maximum allowable information amount per block, and is calculated from a predetermined target buffer occupation amount and a total number of blocks constituting one image frame. The encoding control device according to claim 2, wherein the encoding control device is characterized in that: 5. The predetermined threshold value is a maximum allowable information amount per block. The predetermined threshold value is a predetermined target generated information amount and an effective block (when an image frame is encoded) in the previous image frame. The encoding control device according to claim 1, wherein the encoding control device is calculated from the total number of blocks in which encoding coefficients have been generated. 6. The predetermined threshold value is a maximum allowable information amount per block. The predetermined threshold value is a predetermined target buffer occupancy amount and an effective block in the previous image frame (when an image block is coded). 3. The coding control device according to claim 2, wherein the calculation is performed based on a total number of blocks in which coding coefficients have been generated. 7. The encoding control apparatus according to claim 3, wherein the target generated information amount is corrected according to an effective transmission rate (throughput). 8. The coding control apparatus according to claim 7, wherein the correction of the target generated information amount is performed according to a ratio of (the effective transmission rate / standard transmission rate). 9. The encoding control device according to claim 3, wherein the target generated information amount is corrected according to a state of movement of a subject included in an image, or a state of movement end. 10. The encoding control device according to claim 5, wherein the total number of valid blocks is an average of the total number of valid blocks in a plurality of past image frames. 11. The encoding control device according to claim 6, wherein the total number of valid blocks is an average of the total number of valid blocks in a plurality of past image frames. 12. The block-based quantization control means,
On the basis of the comparison result of the comparing means, the quantization characteristic of the block to be encoded next is determined to be equal to, larger than, or smaller than the reference quantization characteristic. 3. The encoding control device according to claim 1, wherein 13. The block-based quantization control means,
The quantization characteristic of the block to be encoded first in the image frame is set as the reference quantization characteristic, and based on the comparison result of the comparison means, the quantization characteristic of the block to be encoded subsequently is calculated. 3. The encoding control device according to claim 1, wherein the quantization control unit determines the quantization characteristic to be equal to, larger than, or smaller than the previous block. 14. The block-based quantization control means,
3. The encoding control according to claim 1, wherein a quantization characteristic of a block to be encoded next is determined within a preset width centered on the reference quantization characteristic. apparatus. 15. The coding control apparatus according to claim 3, wherein the maximum allowable information amount is updated each time a block is coded according to the following equation. Max_info = (target_info-total_info) / (total_BK-
codec_BK) Max_info: The maximum allowable information amount target_info: The target generated information amount total_info: The generated information amount in the current image frame up to the present time totak_BK: The total number of blocks constituting the image frame codec_BK: The code of the current image frame up to the current time Number of blocks 16. The encoding control apparatus according to claim 4, wherein the maximum allowable information amount is updated each time a block is encoded according to the following equation. Max_info = (target_info-total_info) / (total_BK-
codec_BK) Max_info: the maximum allowable information amount target_info: the target buffer occupancy total_info: the buffer occupancy up to the current time in the current image frame totak_BK: the total number of blocks constituting the image frame codec_BK: the code up to the current time in the current image frame Number of blocks 17. The encoding control apparatus according to claim 5, wherein the maximum allowable information amount is updated each time a block is encoded according to the following equation. Max_info = (target_info-total_info) / (pre_active
_BK-active_BK) Max_info: The maximum allowable information amount target_info: The target generated information amount total_info: The generated information amount in the current image frame up to the present time pre_active_BK: The total number of active blocks in the previous image frame active_BK: The current time in the current image frame Number of the effective blocks up to 18. The coding control apparatus according to claim 6, wherein the maximum allowable information amount is updated each time a block is coded according to the following equation. Max_info = (target_info-total_info) / (pre_active
_BK-active_BK) Max_info: The maximum allowable information amount target_info: The target buffer occupancy total_info: The buffer occupancy up to the current time in the current image frame pre_active_BK: The total number of the effective blocks of the previous image frame active_BK: The current time of the current image frame Number of the effective blocks up to 19. The apparatus according to claim 10, wherein the maximum allowable information amount is updated each time a block is encoded according to the following equation. Max_info = (target_info-total_info) / (ave_active
_BK-active_BK) Max_info: The maximum allowable information amount target_info: The target generated information amount total_info: The generated information amount up to the current time in the current image frame ave_active_BK: The average of the total number of the effective blocks in the past plural image frames active_BK: The current image The number of valid blocks so far in the frame 20. The encoding control apparatus according to claim 11, wherein the maximum allowable information amount is updated each time a block is encoded according to the following equation. Max_info = (target_info-total_info) / (ave_active
_BK-active_BK) Max_info: The maximum allowable information amount target_info: The target buffer occupancy total_info: The buffer occupancy up to the current time in the current image frame ave_active_BK: The average of the total number of the effective blocks in a plurality of past image frames active_BK: The current image The number of valid blocks so far in the frame 21. The block-based quantization control means, when the amount of generated information per block is larger than a second predetermined threshold value, the block-based quantization control unit performs quantization with a quantization characteristic larger than the set quantization characteristic. , Controlling the encoding of the next block, and when the amount of generated information per block is smaller than a third threshold value, the following quantization characteristic is smaller than the set quantization characteristic. The coding control device according to claim 1, wherein coding control of the block is controlled. 22. When the buffer occupancy per block is larger than a predetermined second threshold value, the block unit quantization control means performs a quantization operation with a quantization characteristic larger than the set quantization characteristic. , Controlling the encoding of the next block, and when the buffer occupancy per block is smaller than a third threshold value, the following quantization characteristic is smaller than the set quantization characteristic. 3. The coding control device according to claim 2, wherein coding control of the block is controlled. 23. Each time one image frame is encoded,
A movement start / end detection means for calculating a movement amount of the subject included in the image and detecting one or both of the movement start and the movement end of the subject based on the movement amount; In the detection means, when the movement start state is detected, the total number of blocks of the previous image frame is corrected upward, and when the movement end state is detected,
4. A motion start / end correction means for downwardly correcting the total number of blocks of the previous image frame.
Or the encoding control device according to 4. 24. Each time one image frame is encoded,
A movement start / end detection means for calculating a movement amount of the subject included in the image and detecting one or both of the movement start and the movement end of the subject based on the movement amount; In the detecting means, when the motion start state is detected, the total number of valid blocks of the previous image frame is corrected upward, and when the motion end state is detected, the total number of valid blocks of the previous image frame is corrected downward. Further comprising start / end correction means,
The encoding control device according to claim 5. 25. Each time one image frame is encoded,
A movement start / end detection means for calculating a movement amount of the subject included in the image and detecting one or both of the movement start and the movement end of the subject based on the movement amount; In the detection means, when the motion start state is detected, the average of the total number of valid blocks of the previous image frame is corrected upward, and when the motion end state is detected, the average of the total number of valid blocks of the previous image frame is calculated. The encoding control device according to claim 10, further comprising a movement start / end correction unit that performs downward correction. 26. When the total number of valid blocks of the current image frame exceeds the total number of valid blocks of the previous image frame, a predetermined value is added to the total number of valid blocks of the previous image frame to correct it. The encoding control device according to claim 5, wherein: 27. If the total number of invalid blocks of the current image frame (the block in which a coding coefficient did not occur when the image frame was encoded) exceeds the total number of invalid blocks of the previous image frame, The encoding control device according to claim 5, wherein a predetermined value is corrected by subtracting the value from the total number of effective blocks of the previous image frame. 28. When the total number of valid blocks in the current image frame exceeds an average of the total number of valid blocks in a plurality of past image frames, a predetermined value is set to a value of the total number of valid blocks in the plurality of past image frames. 2. The method according to claim 1, wherein the correction is performed by adding to the average.
12. The encoding control device according to 0 or 11. 29. The total number of invalid blocks (blocks in which no coding coefficient was generated when the image frame was encoded) of the current image frame exceeded the average of the total number of invalid blocks in a plurality of past image frames. The encoding control device according to claim 10, wherein in the case, a predetermined value is corrected by subtracting the predetermined value from an average of the total number of effective blocks in the plurality of past image frames. 30. The method according to claim 30, wherein the amount of motion of the subject is a sum of absolute values of motion vectors output as a result of a motion vector compensation process used when an image frame is subjected to inter-frame predictive coding. An encoding control device according to any one of claims 22 to 24. 31. The coding control apparatus according to claim 1, wherein said block is a macroblock. 32. An encoding control method for controlling encoding of an image frame divided into a plurality of blocks by setting a quantization characteristic for each block. A reference quantization characteristic used for encoding an image frame) is a data amount generated by encoding an image frame immediately before the current image frame (hereinafter, referred to as a previous image frame). Determining based on the amount of generated information; calculating the amount of generated information per block each time one of the blocks constituting the current image frame is encoded; Comparing the calculated amount of generated information per block with the calculated amount of information per block; and encoding based on the result of the comparison and the reference quantization characteristic. The and a step of setting a quantization characteristic used to encode the next block of the block, the coding control method. 33. An encoding control method for controlling encoding of an image frame divided into a plurality of blocks by setting a quantization characteristic for each of the blocks. A reference quantization characteristic used for encoding of an image frame is used for data transmission after encoding of an immediately preceding image frame (hereinafter, referred to as a previous image frame) with respect to the current image frame. Determining based on the data occupancy in the smoothing buffer (hereinafter referred to as buffer occupancy), and each time one of the blocks constituting the current image frame is encoded, the buffer occupancy per block is calculated. Calculating an amount; comparing a predetermined threshold value with the calculated buffer occupancy per block; Based on the quantization characteristic of the result and the basis for comparison, and a step of setting a quantization characteristic used to encode the next block of encoded said block, coding control method. 34. The predetermined threshold value is a maximum allowable information amount per block, and is calculated from a predetermined target generated information amount and a total number of blocks constituting one image frame. 33. The encoding control method according to claim 32, characterized by: 35. The predetermined threshold value is a maximum allowable information amount per block, and is calculated from a predetermined target buffer occupancy amount and a total number of blocks constituting one image frame. The encoding control method according to claim 33, wherein the encoding control method is characterized in that: 36. The predetermined threshold value is a maximum allowable information amount per block, and includes a predetermined target generated information amount and a valid block (when an image frame is coded) in the previous image frame. 33. The coding control method according to claim 32, wherein the calculation is performed based on the total number of blocks in which coding coefficients have been generated. 37. The predetermined threshold value is a maximum allowable information amount per block. The predetermined threshold value is a predetermined target buffer occupancy amount and an effective block (when an image frame is encoded) in the previous image frame. 34. The encoding control method according to claim 33, wherein the calculation is performed from the total number of blocks in which encoding coefficients have been generated. 38. The encoding control method according to claim 34, wherein the target generated information amount is corrected according to an effective transmission rate (throughput). 39. The encoding control method according to claim 38, wherein the correction of the target generated information amount is performed according to a ratio of (the effective transmission rate / standard transmission rate). 40. The encoding control method according to claim 34, wherein the target generated information amount is corrected according to a movement start state or a movement end state of a subject included in the image. 41. The encoding control method according to claim 36, wherein the total number of valid blocks is an average of the total number of valid blocks in a plurality of past image frames. 42. The encoding control method according to claim 37, wherein the total number of valid blocks is an average of the total number of valid blocks in a plurality of past image frames. 43. The step of setting the quantization characteristic, wherein, based on the result of the comparison, the quantization characteristic of a block to be encoded next is equal to or larger than the reference quantization characteristic. 34. The encoding control method according to claim 32, wherein the encoding control method is determined to be any one of a quantization characteristic and a smaller quantization characteristic. 44. The step of setting the quantization characteristic, wherein a quantization characteristic of a block to be encoded first in an image frame is set as the reference quantization characteristic, and based on a result of the comparison, 34. The method according to claim 32, wherein a quantization characteristic of a block to be encoded after the next block is determined to be equal to, larger than, or smaller than that of the previous block. Coding control method according to any one of the preceding claims. 45. The step of setting the quantization characteristic includes determining a quantization characteristic of a block to be encoded next within a preset width centered on the reference quantization characteristic. The encoding control method according to claim 32, wherein the encoding control method is characterized by: 46. The encoding control method according to claim 34, wherein the maximum allowable information amount is updated each time a block is encoded according to the following equation. Max_info = (target_info-total_info) / (total_BK-
codec_BK) Max_info: The maximum allowable information amount target_info: The target generated information amount total_info: The generated information amount in the current image frame up to the present time totak_BK: The total number of blocks constituting the image frame codec_BK: The code of the current image frame up to the current time Number of blocks 47. The encoding control method according to claim 35, wherein the maximum allowable information amount is updated each time a block is encoded according to the following equation. Max_info = (target_info-total_info) / (total_BK-
codec_BK) Max_info: the maximum allowable information amount target_info: the target buffer occupancy total_info: the buffer occupancy up to the current time in the current image frame totak_BK: the total number of blocks constituting the image frame codec_BK: the code up to the current time in the current image frame Number of blocks 48. The encoding control method according to claim 36, wherein the maximum allowable information amount is updated every time a block is encoded according to the following equation. Max_info = (target_info-total_info) / (pre_active
_BK-active_BK) Max_info: The maximum allowable information amount target_info: The target generated information amount total_info: The generated information amount in the current image frame up to the present time pre_active_BK: The total number of active blocks in the previous image frame active_BK: The current time in the current image frame Number of the effective blocks up to 49. The encoding control method according to claim 37, wherein the maximum allowable information amount is updated every time a block is encoded according to the following equation. Max_info = (target_info-total_info) / (pre_active
_BK-active_BK) Max_info: The maximum allowable information amount target_info: The target buffer occupancy total_info: The buffer occupancy up to the current time in the current image frame pre_active_BK: The total number of the effective blocks of the previous image frame active_BK: The current time of the current image frame Number of the effective blocks up to 50. The encoding control method according to claim 41, wherein the maximum allowable information amount is updated each time a block is encoded according to the following equation. Max_info = (target_info-total_info) / (ave_active
_BK-active_BK) Max_info: The maximum allowable information amount target_info: The target generated information amount total_info: The generated information amount up to the current time in the current image frame ave_active_BK: The average of the total number of the effective blocks in the past plural image frames active_BK: The current image The number of valid blocks so far in the frame 51. The encoding control method according to claim 42, wherein the maximum allowable information amount is updated each time a block is encoded according to the following equation. Max_info = (target_info-total_info) / (ave_active
_BK-active_BK) Max_info: The maximum allowable information amount target_info: The target buffer occupancy total_info: The buffer occupancy up to the current time in the current image frame ave_active_BK: The average of the total number of the effective blocks in a plurality of past image frames active_BK: The current image The number of valid blocks so far in the frame 52. The step of setting the quantization characteristic, wherein, when the amount of generated information per block is larger than a second predetermined threshold value, the quantization characteristic is larger than the set quantization characteristic. Then, the encoding of the next block is controlled, and when the amount of generated information per block is smaller than a predetermined third threshold, the next block has a quantization characteristic smaller than the set quantization characteristic. 33. The encoding control method according to claim 32, wherein encoding of the block is controlled. 53. The step of setting the quantization characteristic, wherein when the buffer occupancy per block is larger than a second predetermined threshold, the quantization characteristic is larger than the set quantization characteristic. Then, the encoding of the next block is controlled, and if the buffer occupancy per block is smaller than a third threshold value, the next quantization characteristic is smaller than the set quantization characteristic. 34. The encoding control method according to claim 33, wherein encoding of the block is controlled. 54. Each time one image frame is encoded,
Calculating the amount of movement of the subject included in the image, detecting one or both of the start and end states of the movement of the subject based on the amount of movement; and Correcting the total number of blocks of the previous image frame upward if detected, and correcting the total number of blocks of the previous image frame downward if an end-of-motion condition is detected. Encoding control method. 55. Each time one image frame is encoded,
Calculating the amount of movement of the subject included in the image, detecting one or both of the start and end states of the movement of the subject based on the amount of movement; and Correcting the total number of valid blocks of the previous image frame upward if detected, and correcting the total number of valid blocks of the previous image frame downward if an end-of-motion condition is detected. Or the encoding control method according to 37. 56. Each time one image frame is encoded,
Calculating the amount of movement of the subject included in the image, detecting one or both of the start and end states of the movement of the subject based on the amount of movement; and If detected, the average of the total number of valid blocks of the previous image frame is corrected upward, and if the motion end state is detected, the average of the total number of valid blocks of the previous image frame is corrected downward. Claim 41
Or the encoding control method according to 42. 57. When the total number of valid blocks of the current image frame exceeds the total number of valid blocks of the previous image frame, a predetermined value is added to the total number of valid blocks of the previous image frame to make correction. The encoding control method according to claim 36 or 37, wherein: 58. When the total number of invalid blocks of the current image frame (blocks in which a coding coefficient is not generated when the image frame is coded) exceeds the total number of invalid blocks of the previous image frame, 38. The encoding control method according to claim 36, wherein a predetermined value is corrected by subtracting the value from the total number of effective blocks of the previous image frame. 59. When the total number of valid blocks in the current image frame exceeds the average of the total number of valid blocks in a plurality of past image frames, a predetermined value is set to a value of the total number of valid blocks in the past plurality of image frames. 5. The correction in addition to the average.
43. The encoding control method according to 1 or 42. 60. The total number of invalid blocks (blocks in which a coding coefficient did not occur when the image frame was encoded) of the current image frame exceeded the average of the total number of invalid blocks in a plurality of past image frames. 43. The encoding control method according to claim 41, wherein in the case, a predetermined value is corrected by subtracting from an average of the total number of the effective blocks in the past plurality of image frames. 61. The method according to claim 61, wherein the motion amount of the subject is a sum of absolute values of motion vectors output as a result of a motion vector compensation process used when interframe predictive coding of an image frame is performed. The encoding control method according to any one of claims 54 to 56. 62. The encoding control method according to claim 32, wherein said block is a macroblock. 63. A medium recorded with an encoding control method for controlling encoding of an image frame divided into a plurality of blocks by setting a quantization characteristic for each block as a program executable by a computer device A quantization characteristic that is used as a reference for encoding an image frame to be processed (hereinafter, referred to as a current image frame) is set to the image frame immediately before the current image frame (hereinafter, a previous image frame). (A frame) is determined based on the amount of generated information, which is the amount of data generated by the encoding of the current image frame. Each time one of the blocks constituting the current image frame is encoded, Calculating an information amount; comparing a predetermined threshold value with the calculated generated information amount per block; Setting a quantization characteristic used for encoding a block next to the encoded block based on the result of the comparison and the reference quantization characteristic. Recording medium. 64. A medium in which an encoding control method for controlling encoding of an image frame divided into a plurality of blocks by setting a quantization characteristic for each block is recorded as a program executable by a computer device. A quantization characteristic that is used as a reference for encoding an image frame to be processed (hereinafter, referred to as a current image frame) is set to the image frame immediately before the current image frame (hereinafter, a previous image frame). Determining on the basis of the data occupancy (hereinafter referred to as buffer occupancy) in the smoothing buffer used for the data transmission after encoding of the current image frame. Calculating a buffer occupancy per block each time encoding is performed; a predetermined threshold value; Comparing the buffer occupancy per block with a quantization characteristic used for encoding a block next to the encoded block based on the comparison result and the reference quantization characteristic. And a step of setting at least a program for executing the program.