JPH0514873A - Image encoder - Google Patents

Abstract

PURPOSE:To improve encoding efficiency by obviating a circuit, etc., detecting a memory and a difference for movement detection to make a circuit scale small and detecting movement in parallel with DCT to attain the shortening of a processing time and efficient movement detection. CONSTITUTION:A blocking circuit 12 divides image data into a block setting nXn pieces in spatial arrangement to be one block. A DCT circuit 13 calculates a conversion coefficient by DCT of image data in the respective blocks. A quantization circuit 14 sorts the respective blocks into moving images and still images based on the conversion coefficient, sets the area of low frequency components to be an encoding area at a still image area, sets the areas of low/high frequent components to be an encoding area at a moving image block, forms quantization data by quantizing the conversion coefficiency of the encoding area and outputs the quantization data.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an image coding apparatus, and more particularly to an image coding apparatus for highly efficient coding image data by discrete cosine transform.

[0002]

2. Description of the Related Art When transmitting image data or recording it on a recording medium such as a magnetic tape, various encodings are used for compressing image information. For example, so-called predictive coding, transform coding, vector quantization, etc. are known.

By the way, the transform coding utilizes the correlation of image signals to transform sample values (hereinafter referred to as image data) into mutually orthogonal axes to uncorrelate the correlation between image data, The so-called basis vectors are orthogonal to each other, the sum of the average signal power before conversion and the sum of the average power of so-called conversion coefficients obtained by orthogonal conversion are equal, and the power is concentrated on low-frequency components. An excellent orthogonal transform is adopted, for example, so-called Hadamard transform, Haar transform, Karnen-Roube (KL) transform,
Discrete Cosine Transfor
m), discrete sine transformation (hereinafter DST: Discrete Sin)
e Transform), slant transform, etc. are known.

Here, the DCT will be briefly described. The DCT divides an image into image blocks each consisting of n (n × n) pixels in the horizontal and vertical directions in a spatial arrangement, and orthogonally transforms image data in the image blocks using a cosine function. The DCT has come to be widely used for transmission and recording of image data because a high-speed operation algorithm exists and a one-chip so-called LSI that enables real-time conversion of image data is realized.
Further, the DCT has almost the same characteristics as the KL conversion, which is the optimum conversion in terms of the coding efficiency in terms of the degree of power concentration on the low-frequency component that directly affects the efficiency. Therefore, it is possible to significantly reduce the information amount as a whole by encoding the transform coefficient obtained by the DCT only for the component where the power is concentrated.

Specifically, n × n image data is DC
The conversion coefficient obtained by T is, for example, C _ij (i = 0 to n−
1, j = 0 to n−1), the conversion coefficient C ₀₀ corresponds to the DC component representing the average luminance value in the image block, and its power is usually considerably larger than the other components. Therefore, when the coarsely quantizing the DC components, from where the so-called block distortion is visually orthogonal transform coding specific noise felt as a large image quality degradation occurs, the number of number of bits to transform coefficients C ₀₀ (e.g., 8 (Equal to or more bits) are equally quantized, and the conversion coefficient C _ij (excluding C ₀₀ ) of the other components (hereinafter referred to as AC components) excluding the DC component is, for example, that the visual spatial frequency decreases in the high range. By utilizing the characteristics, the higher the frequency component is, the more the bit number is allocated and the quantization is performed.

Further, for example, in a region corresponding to one image block of transform coefficients C _ij, to detect the maximum line number and column number value exists significant transform coefficients C _ij other than zero (≠ 0), a significant _So- called zone coding in which transform coefficients are surrounded by a rectangle determined by the rows and columns, and only the transform coefficients C _ij included in the enclosed area (hereinafter, such an area is simply referred to as a coding area) are quantized. In (Zonal Coding), an image block is classified into a moving image block with motion and a still image block without motion, and quantization is performed by changing the size of the coding area between the moving image block and the still image block to improve coding efficiency. Is being raised.

[0007] Specifically, in the still image block, for example, as shown in FIG. 8, when the size of the transform coefficients C _ij corresponds to an image block area 80, for example, 8 × 8, the low significant transform coefficients C _ij Focus on the frequency domain. On the other hand, in the moving image block, as shown in FIG. 9, for example, the degree of concentration of the significant conversion coefficient C _{ij in} the low frequency region is low.

Therefore, the conventional apparatus is provided with two so-called frame memories, detects a difference between frames for each image block, sets an image block having a large difference as a moving image block, and an image block having a small difference as a still image block. Then, the image blocks are classified. And
By increasing the coding area for moving picture blocks and narrowing the coding area for still picture blocks and adaptively quantizing, the amount of data is reduced and coding efficiency is improved. It has become.

In the transmission and recording of image data, the transform coefficient C _ij obtained by DCT of the image data is quantized as described above, and so-called Huffman coding or Run length coding (Ru
Variable length coding such as n length coding) is performed, and a sync signal and a parity are added to the obtained coded data for transmission and recording.

Further, in a digital video tape recorder (hereinafter simply referred to as VTR) for recording a video signal as a digital signal on a magnetic tape, the data amount of one frame or one field is fixed (fixed length) in consideration of editing and variable speed reproduction. In view of the circuit scale, it is desirable that the data amount of encoded data of one image block or the data amount of a processing unit in which the encoded data is collected by a predetermined number of image blocks is constant. Therefore, in the VTR, a plurality of quantizers having different quantizing widths are prepared, and for example, a quantizer having a data amount of an image block equal to or less than a predetermined value and having a smallest quantizing width is selected and quantized. Is supposed to do.

[0011]

By the way, although high coding efficiency can be obtained by the adaptive quantization accompanied by the motion detection in the zone coding described above, it is possible to obtain the two coding values as described above for the motion detection. A frame memory, a circuit for detecting a difference, and the like are required, and the circuit scale is large in the conventional device. Further, the processing time is long due to the motion detection.

The present invention has been made in view of the above circumstances, and it is possible to reduce the circuit scale, the processing time, and the motion detection as compared with the conventional device. It is an object of the present invention to provide an image coding apparatus which can perform the above-mentioned processing with high accuracy and has high coding efficiency.

[0013]

According to the present invention, in order to solve the above-mentioned problems, image data is arranged in a spatial arrangement of n × n.
Blocking means for dividing each block into one block, discrete cosine transforming means for orthogonally transforming image data of each block from the blocking means using a cosine function to calculate transform coefficients, and the discrete cosine A moving image in which each block moves based on the quantizing means that quantizes the transform coefficient from the transforming means to form quantized data and outputs the quantized data, and the transform coefficient from the discrete cosine transforming means. Detects whether the block is a still image block with no motion, quantizes the low frequency component of the transform coefficient when the block is a still image block, and quantizes the low frequency component and the high frequency component of the transform coefficient when the block is a moving image block. Control means for controlling the quantizing means so as to change the value.

[0014]

In the image coding apparatus according to the present invention, the image data is divided into blocks in which n × n in the spatial arrangement is one block, and the image data of each block is orthogonally transformed using the cosine function to obtain transform coefficients. Is calculated, and the transform coefficient is quantized to form quantized data, and when the quantized data is output, whether each block is a moving image block with motion or a still image block with no motion is determined based on the transform coefficient. When the block is a still image block, the low frequency component of the transform coefficient is quantized, and when the block is a moving image block, the low frequency component and the high frequency component of the transform coefficient are quantized.

[0015]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment of an image coding apparatus according to the present invention will be described below with reference to the drawings. FIG. 1 shows a circuit configuration of an image coding apparatus to which the present invention is applied, and FIG. 2 is a circuit configuration of a recording system of a digital video tape recorder (hereinafter simply referred to as VTR) to which the image coding apparatus is applied. FIG. 3 shows a circuit configuration of a reproduction system of the VTR.

First, the VTR will be described. As shown in FIG. 2, this VTR converts an analog video signal into a digital signal, performs data processing such as so-called conversion coding on the obtained image data to perform data compression, and then, through the magnetic head 21. A recording system for recording on the magnetic tape 1 and the magnetic tape 1 to the magnetic head 3 as shown in FIG.
The reproduction signal reproduced by 1 is binarized, and after being subjected to data processing such as decoding, it is converted into an analog signal and reproduced as an analog video signal.

The recording system, as shown in FIG.
An analog / digital converter (hereinafter referred to as an A / D converter) 11 that samples a video signal and converts it into a digital signal to form image data, and the image data from the A / D converter 11 is n in a spatial arrangement. A predetermined number of image blocks are divided into image blocks G _h (h = 1 to H, H is dependent on the number of pixels in one frame and the number of pixels n ² of the image blocks G _h ) in which n is one block. Blocking circuit 12 that forms a processing unit, which is one unit of data processing or transmission, that is composed of G _h , and the blocking circuit 1
The image data from 2 is orthogonally transformed (hereinafter referred to as DCT: Discrete Cosine Transform) using the cosine function, and the transformation coefficient C _ij (i = 0 to n−1, j =) of each image block G _h
Discrete cosine transform circuit (hereinafter DCT) for calculating 0 to n-1)
Circuit 13 and a quantizing circuit 1 for quantizing transform coefficients C _ij from the DCT circuit 13 to form quantized data.
4 and the quantized data from the quantization circuit 14 are encoded by, for example, a so-called variable length code to obtain encoded data VLC _ij.
Encoding circuit 15 forming (i = 0 to n-1, j = 0 to n-1), and encoded data VL from the encoding circuit 15
For example, a parity addition circuit 17 for adding a parity for error detection or error correction to C _ij for each processing unit, and a coded data VLC _ij to which a parity from the parity addition circuit 17 is added are provided with a synchronization signal and an image. An identification bit (hereinafter referred to as an ID) for identifying the number h or the like of the block G _h is added to each processing unit to form transmission data, and a synchronization signal insertion circuit 18 sends parallel data from the synchronization signal insertion circuit 18. Parallel / serial (hereinafter referred to as P / S) converter 1 for converting incoming transmission data into serial data
9 and a channel encoder (hereinafter referred to as ENC) 20 which applies a modulation process suitable for recording to the transmission data from the P / S converter 19 to generate a recording signal and supplies the recording signal to the magnetic head 21. ..

In this recording system, after converting a video signal supplied as an analog signal via the terminal 2 into image data, for example, one frame of image data is divided into image blocks G _h , and each image block is divided into image blocks G _h. The transform coefficient C _ij is calculated by performing DCT on the image data of G _h, the transform coefficient C _ij is quantized to form quantized data, and the quantized data is encoded by the variable length code to obtain the encoded data VLC _ij.
Are formed. Further, in this recording system, a sync signal or the like is added to the encoded data VLC _ij for each processing unit to form transmission data, and then the transmission data is subjected to modulation suitable for recording, for example, scrambling or NRZI modulation processing, Recording is performed on the magnetic tape 1 by the magnetic head 21.

Thus, the image coding apparatus according to the present invention,
That is, the main part of the VTR configured as described above is composed of the blocking circuit 12 to the quantizing circuit 14, and is specifically as follows.

As shown in FIG. 1, for example, the blocking circuit 12 has a memory capacity of, for example, one frame, stores a memory 12a for storing image data, and n × n the image data from the memory 12a in a spatial arrangement. with dividing n number of the image blocks G _h to 1 block, composed of a blocking unit 12b read out for each processing unit consisting of the image block G _h predetermined number obtained by dividing one frame into a plurality.

The blocking circuit 12 stores the image data supplied through the terminal 4 in the memory 12a for each frame, and the image data stored in the memory 12a in the spatial arrangement of, for example, 8 bits. The image block G _h is divided into 8 × 8 blocks as one block, and the read image data is supplied to the DCT circuit 13 for each processing unit consisting of a predetermined number of image blocks G _h .

The DCT circuit 13 is, for example, a so-called DSP.
(Digital Signal Processor) and the like, and is composed of a DCT 13a that performs orthogonal transform in the vertical direction using a cosine function and a DCT 13b that performs orthogonal transform in the horizontal direction (hereinafter, the orthogonal transform in one direction is referred to as a one-dimensional DCT, The orthogonal transformation in two directions is referred to as a two-dimensional DCT), the block circuit 12
2D DCT of image data supplied from
Then, the transform coefficient C _ij is calculated, and the transform coefficient C _ij is supplied to the quantization circuit 14.

As shown in FIG. 1, the quantization circuit 14 includes a buffer memory 14a for storing the conversion coefficient C _ij from the DCT circuit 13 for each processing unit, and the conversion read from the buffer memory 14a. A quantizer Q ₁ , which quantizes the coefficients C _ij to form quantized data,
The motion of each image block G _h is detected based on Q ₂ and Q ₃ and the conversion coefficient in the vertical direction from the DCT 13a, and these image blocks G _h are made into a moving image block with motion and a still image block without motion. Motion detection circuit 50 for classification
Based on the classification information from the motion detection circuit 50, among the conversion coefficients C _ij from the DCT circuit 13, significant conversion coefficients C _ij (≠ 0) other than zero are set as moving image blocks and still image blocks. In order to perform so-called zone coding in which only the transform coefficients C _ij of the enclosed area are quantized by being surrounded by predetermined shapes different from each other, the area to be quantized (hereinafter referred to as an encoding area) Of the image block G _h after the quantization of the transform coefficient C _ij included in the coding region decided by the region decision circuit 14b for deciding the image block G _h by the so-called Huffman code. and detects the amount of data, and the code amount calculation circuit 14c for controlling the quantization width of the quantizer Q _1, Q _2, Q ₃ based on the amount of data, said code amount calculation circuit 1
Huffman coding rule in 4c (hereinafter referred to as table)
A Huffman code table circuit 14d to give, the code amount calculation circuit 14c based on the quantizer Q _1, Q control of
_2, composed of a selector 14e for selecting one of the outputs of the Q _3.

Further, the motion detecting circuit 50 is, for example, as shown in FIG. 4, an accumulating circuit for accumulating n columns of vertical transform coefficients supplied from the DCT 13a through a terminal 55 in order from a high frequency component. 51 and the accumulating circuit 51
Based on the detected row number k from the comparison circuit 52, which compares the cumulative value from the predetermined threshold value TH with the predetermined threshold value TH to detect the row number k whose cumulative value exceeds the threshold value TH. It is composed of a classification circuit 53 for classifying the block G _h into a moving image block and a still image block.

Then, the quantizing circuit 14 zone-codes the transform coefficient C _ij from the DCT circuit 13 by using coding regions having different shapes for the moving image block and the still image block, and at the same time, for the image block G _h . The amount of data is equal to or less than a predetermined amount of data, and quantization data is quantized (finely) with a minimum quantization width so that quantization distortion is minimized, and the quantized data is supplied to the encoding circuit 15. It is supposed to do.

More specifically, the accumulator circuit 51 includes a DCT 13
Transform coefficients D _ij (i = 0 to n-1, j = 0 to n-1) obtained by performing one-dimensional DCT on the image data from a in the vertical direction
For example, the absolute value is sequentially accumulated from the high frequency component in the vertical direction, and the accumulated value is output for each row. For example, image block G
_{Assuming that} the size of _h is 8 × 8, the total T _i of the conversion coefficients D _ij of each row can be calculated by the following equation (1).

T _i = | D _i0 | + | D _i1 | + | D _i1 | + ... + | D _i7 | (1)

Then, the total T of the conversion coefficients D _ij of each row
_i is sequentially accumulated from high frequency components as shown in the following formula (2), and the accumulated value S is supplied to the comparison circuit 52 for each row. S _i = S _{i + 1} + T _i (i = 7 to 0) (2)

The comparison circuit 52 uses the accumulated value S _i and the threshold value T _i.
By comparing H for each row, the row number k for which the cumulative value S _i exceeds the threshold value TH is detected, and this row number k is supplied to the classification circuit 53.

The classification circuit 53 sets the row number k to a predetermined value,
For example, when compared with “5” and the line number k is “5” or more, the image block G _h is regarded as the moving image block and the line number k.
Is less than “5”, the image block G _h is determined as a still image block. Then, for example, logic "1" in the moving image block and logic "0" in the still image block are used as classification information through the terminal 56 to determine the area determination circuit 14b shown in FIG. 1 and the parity addition circuit shown in FIG. 17
Supply to.

Thus, the motion detection circuit 50 has the DCT1
Based on the distribution of the transform coefficients D _ij obtained by one-dimensional DCT in the vertical direction from 3a, the image block G _h containing a large amount of high frequency components is set as a moving image block, and the image block G _h containing a small amount of high frequency components is _defined as Each image block G _h is classified as a still image block, and the classification information is used as the area determination circuit 1.
4b.

The area determination circuit 14b is used by the image block G _h.
Is a still image block, for example, as shown in FIG. 5A, in the region 60 corresponding to one image block G _h of the transform coefficient C _ij from the DCT circuit 13, the significant transform coefficient C _ij (value of which is non-zero). ≠ 0) is surrounded by a quadrangle, and the enclosed area (shown by diagonal lines) is a coding area 61. Further, for example, as shown in FIG. 5b, the significant conversion coefficient C _ij is enclosed by a triangle, and this enclosed area is surrounded by a triangle. The coding area 62 is used. on the other hand,
When the image block G _h is a moving image block, the area determination circuit 14 b encodes an area including a high frequency component in the vertical direction in addition to the above-described encoded area 61 having a rectangular shape as shown in FIG. 6A. The area 63 is used, and, for example, FIG.
As shown in FIG. 5b, the above-described shape is a triangular encoded area 62.
In addition to the above, a region including a high frequency component in the vertical direction is set as a coding region 64. Then, the area determination circuit 14b uses the area information for identifying the encoded area, for example, in the still image block, the row number U ₁ and the column number V ₁ shown in FIG. 5a, or the row number shown in FIG. 5b. U ₂ and the column number V ₂ are supplied to the code amount calculation circuit 14c, while in the moving image block,
The column numbers v ₁ and v ₂ shown in FIG. 6a or the row numbers u ₁ and column numbers v ₂ and v ₃ shown in FIG. 6b are supplied to the code amount calculation circuit 14c.

The code amount calculating circuit 14c is a region determining circuit 1
Based on the area information from 4b, the image block G is formed by combining a plurality of quantization widths such that the transform coefficient C _ij included in the coding area is quantized with a smaller quantization width for lower frequency components.
After quantization for each _h , Huffman code table circuit 14d
To detect the data amount of the image block G _h by performing encoding based on the Huffman code table from (1) to (4) and the data amount of the image block G _h is a predetermined value or less and the minimum quantization width (quantization distortion becomes minimum. ) Detect the combination. Then, the code amount calculation circuit 14c supplies the quantizers Q ₁ , Q ₂ , and Q ₃ with the quantization number m indicating the combination of the detected quantization widths, and supplies the region information to the selector 14e. The quantization number m and the area information are supplied to the parity adding circuit 17 shown in FIG. 2 via the terminals 6 and 7, respectively.

The quantizers Q ₁ , Q ₂ and Q ₃ have their respective quantization widths controlled on the basis of the quantization number m from the code amount calculating circuit 14c. For example, the quantization number m Is “0” (hereinafter simply referred to as m = 0), all of the quantizers Q ₁ , Q ₂ , and Q ₃ perform quantization with a predetermined quantization width q, and when m = 1, the quantizer Q ₁ and Q ₂ quantize with a quantization width q, quantizer Q ₃ quantizes with a quantization width 2q, and when m = 2, quantizer Q ₁ quantizes with a quantization width q. And quantizers Q ₂ , Q ₃ quantize with a quantization width of 2q, and when m = 3, quantizers Q ₁ , Q ₂ , Q ₃
All perform quantization with a quantization width of 2q ... That is, a quantizer with a smaller quantization number m and a smaller subscript number has a smaller quantization width (finer). Therefore, the transform coefficients C _ij read from the buffer memory 14a are quantized, quantized data of different data amounts are formed for the same transform coefficient C _ij , and the quantized data are selected. 14e
Supply to. In addition, instead of doubling the quantization width (for example, 2q with respect to q), the conversion coefficient C _ij itself may be halved.

The selector 14e includes a code amount calculation circuit 14c.
The quantized data from the quantizers Q ₁ , Q ₂ and Q ₃ are switched and selected based on the area information from For example, as shown in FIG. 7, the selector 14e causes the area 70 corresponding to one image block G _h of the conversion coefficient C _ij to be the area 7 of the low frequency component.
1, intermediate frequency component region 72, high frequency component region 7
When the quantized data is divided into three regions of 3 and corresponds to the transform coefficient C _ij included in the region 71 of the coding region, the quantized data from the quantizer Q ₁ is selected and quantized. When the data corresponds to the transform coefficient C _ij included in the area 72 of the coding area, the quantized data from the quantizer Q ₂ is selected, and the quantized data is the area of the coding area. 73 corresponds to the transform coefficient C _ij , the quantized data from the quantizer Q ₃ is selected, and when the quantized data corresponds to the transform coefficient C _ij of the area other than the coding area, Also, the quantized data from the quantizers Q ₁ , Q ₂ , and Q ₃ are not selected, and the quantized data thus selected is supplied to the encoding circuit 15.

As a result, for example, as shown in FIG. 7 described above, the image block G _h is a still image block and the coding area is the coding area 62 shown in FIG. 5b (shown by a broken line in FIG. 7). Yes, when m = 1, the transform coefficient C _ij included in the region 71 of the coding region 62 is quantized with the quantization width q, and the transform coefficient C included in the region 72 of the coding region 62. Quantized data obtained by quantizing _ij with a quantization width of 2q is supplied to the encoding circuit 15. That is,
The image block G _h is classified into a moving image block and a still image block, the shape of the coding region is changed between the moving image block and the still image block, and in this coding region, the lower the frequency component, the smaller the quantization width (fine) the quantum. The quantized data obtained by the coding is supplied to the coding circuit 15.

The coding circuit 15 is, for example, a Huffman coder for performing variable length coding and a so-called run length code (Run).
Length code) device and the like, and this encoding circuit 15
Encodes the quantized data from the selector 14e with a Huffman code and a run length code, respectively, to form encoded data VLC _ij . The encoded data VLC _ij is added via the terminal 5 to the parity addition shown in FIG. It is adapted to be supplied to the circuit 17.

As a result, the coding circuit 15 outputs coded data VLC _ij obtained by being quantized so that the data amount of the image block G _h is a predetermined value or less and the quantization distortion is minimized. It Then, at this time, the coding area is not uniquely determined as a rectangular area surrounded by the maximum rows and columns in which the significant conversion coefficient C _ij exists as in the conventional apparatus, but one-dimensional in the vertical direction. The image block G _h is classified into a moving image block and a still image block based on the transform coefficient D _ij obtained by DCT, and in the moving image block and the still image block, for example, as shown in FIGS. By setting the areas of different shapes,
For example, the distribution of the conversion coefficient C _ij of the still image block shown in FIG. 8 as described in the related art, that is, the distribution having a high low-frequency concentration, the distribution of the conversion coefficient C _ij of the moving image block shown in FIG. 9,
That is, it is possible to perform quantization that is suitable for each distribution having a low low-frequency concentration, and it is possible to obtain high coding efficiency. Further, the motion detection for classifying the image block G _h into the moving image block and the still image block is performed by using the conversion coefficient D _ij from the DCT 13a, and the frame memory for motion detection and the like as in the conventional device are used. The circuit scale can be further reduced without requiring a circuit for detecting the difference. In addition, since the motion detection is performed based on the conversion coefficient D _ij obtained by the one-dimensional DCT in the vertical direction, that is, based on the frequency analysis in the vertical direction, the accuracy of the motion detection is good. Furthermore, the motion detection can be performed simultaneously with the one-dimensional DCT in the horizontal direction, and the processing time can be shortened.

Then, the parity addition circuit 17 and the synchronization signal insertion circuit 18 shown in FIG. 2 described above include the encoded data VLC _ij from the encoding circuit 15 and the classification information of each image block G _h from the quantization circuit 14, The adopted quantization number m and area information (row numbers u ₁ , U ₁ , column numbers v ₁ , V _1, etc.) are time-division multiplexed, and parity and synchronization signals are added to form transmission data. As a result, for example, one processing unit in order from the beginning is the synchronization signal, the ID, the classification information of each image block G _h , the quantization number m and the region information, the encoded data VLC _{ij of} a predetermined number of image blocks G _h , and the parity. The transmission data is output. Incidentally, the area information as described above is recorded for each image block G _h is in order to be able to recognize the coding region when the regeneration as described below, for example, an image block G _h Size of 8 × 8
Then, since the values of the row numbers u ₁ , U ₁ , column numbers v ₁ , V _1, etc. are “7” at maximum, the number of bits newly required for the area information is 6 (= 3 ×). 2) or 9 (= 3
× 3) bits, which hardly affects the coding efficiency.

As described above, this image coding apparatus temporarily stores the image data supplied via the terminal 4 in the memory 12a, and stores the stored image data in the n spatial arrangement.
The image data is divided into image blocks G _h having × n blocks as one block, and read for each processing unit including a predetermined number of image blocks G _h, and the image data of each image block G _h is DCT.
After that, the obtained transform coefficient C _ij is quantized with a combination of quantization widths in which the data amount of the image block G _h is a predetermined value or less and the quantization distortion is minimum, and this quantized data is variable-length coded. , When outputting the obtained encoded data VLC _ij via the terminal 5, it is 1 in the vertical direction from the DCT 13a.
Based on the transformation coefficient D _ij obtained by the dimension DCT, each image block G _h is classified into a moving image block and a still image block, and the coding region is a low frequency component region in the still image block and low in the moving image block. Encoding efficiency can be increased by quantizing only the transform coefficient C _ij of the encoding region as the region of the frequency component and the high frequency component. In addition, since the motion detection is performed based on the conversion coefficient D _ij obtained by the one-dimensional DCT in the vertical direction, that is, based on the frequency analysis in the vertical direction, the accuracy of the motion detection is good. Furthermore, the motion detection can be performed simultaneously with the one-dimensional DCT in the horizontal direction, and the processing time can be shortened.

Next, the reproducing system of this VTR will be described. As shown in FIG. 3 described above, this reproducing system is a channel decoder (hereinafter simply referred to as DEC) that reproduces transmission data by performing signal processing such as NRZI demodulation on a reproduced signal reproduced from the magnetic tape 1 by the magnetic head 31. 32
And a serial / parallel (hereinafter referred to as S / P) converter 33 for converting transmission data sent as serial data from the DEC 32 into parallel data, and pulling in synchronization of transmission data from the S / P converter 33. Together with the synchronization signal detection circuit 34 for reproducing the encoded data VLC _ij
And a time axis correction circuit (hereinafter TBC: Time B) for correcting the fluctuation of the time axis that occurs during reproduction of the encoded data VLC _ij.
ASE Corrector) 35, an error correction circuit 36 that performs error correction of the encoded data VLC _ij from the TBC 35, and sets an error flag EF for the encoded data VLC _ij that could not be error-corrected. A decoding circuit 37 that decodes the encoded data VLC _ij that has been variable-length encoded at the time of recording from the correction circuit 36 and reproduces quantized data, and dequantization into quantized data from the decoding circuit 37. an inverse quantization circuit 38 to reproduce the transform coefficients C _ij performs signal processing, inverse discrete cosine transform circuit for reproducing image data by orthogonal transformation transform coefficients C _ij from inverse quantization circuit 38 (hereinafter IDCT 39) and the I
A deblocking circuit 40 that forms one frame of image data from the image data supplied from the DCT circuit 39 for each image block G _h , and the deblocking circuit based on the error flag EF from the error correction circuit 36. Error correction circuit 41 for performing error correction on image data from 40
And a digital / analog converter (hereinafter referred to as D / A converter) 42 which converts the image data from the error correction circuit 41 into an analog signal and outputs the analog signal.

Next, the operation of the reproducing system configured as described above will be described. The DEC 32 binarizes the reproduction signal reproduced by the magnetic head 31 from the magnetic tape 1, then performs NRZI demodulation, performs descrambling processing to reproduce the transmission data, and reproduces the transmission data by S / S.
The signal is supplied to the sync signal detection circuit 34 via the P converter 33.

The synchronizing signal detection circuit 34 is used for the S / P converter 3
The sync signal is detected from the transmission data converted into parallel data in step 3 to pull in synchronization, and the encoded data VL
C _ij is reproduced and this encoded data VLC _ij is converted to TBC35.
Supply to.

The TBC 35 corrects the time axis of the encoded data VLC _ij , absorbs the fluctuation of the time axis that occurs during reproduction, and supplies the time axis corrected encoded data VLC _ij to the error correction circuit 36. ..

The error correction circuit 36 uses the encoded data VL.
The error correction of C _ij is performed using the parity added at the time of recording, and the error flag EF is set for the coded data VLC _ij having an error exceeding the error correction capability, and the error-corrected coding is performed. The data VLC _ij is supplied to the decoding circuit 37.

The decoding circuit 37 decodes the coded data VLC _ij coded by the Huffman code and the run length code at the time of recording to reproduce the quantized data, and dequantizes the quantized data. 38.

The dequantization circuit 38 uses the encoded data VLC.
classification information of each image block G _h reproduced together with _ij , quantization number m, and area information (row number u, U, column number v,
Based on V), the combination of the quantization widths used at the time of recording and the coding area are recognized. Then, for example, when the image block G _h is a still image block, the coding region is the coding region 62 shown in FIG. 5b, and when m = 1,
The quantized data included in the region 71 and the encoded region 62 is inversely quantized with the quantization width q, and the quantized data included in the region 72 and the encoded region 62 is inversely quantized with the quantization width 2q. In the areas other than the coding area 62, the conversion coefficient C _ij is reproduced by setting the value to “0”, and this conversion coefficient C _ij is IDCT.
Supply to the circuit 39.

The IDCT circuit 39 orthogonally transforms the transform coefficient C _ij using the transposed matrix corresponding to the transform matrix used at the time of recording, reproduces the image data for each image block G _h, and reproduces this image data. It is supplied to the inverse blocking circuit 40.

The deblocking circuit 40 uses the image block G.
Image data for one frame is formed from the image data reproduced for each _h and supplied to the error correction circuit 41.

The error correction circuit 41 performs an interpolation process using, for example, image data having no error in the vicinity of the image data that cannot be error-corrected by the error correction circuit 36 described above, so that the error-corrected image data cannot be corrected. Error correction is performed, and the image data with this error corrected is supplied to the D / A converter 42.

The D / A converter 42 converts the error-corrected image data into an analog signal, and outputs the analog video signal from the terminal 3 as a luminance signal Y and color difference signals U and V, for example.

As described above, the data amount of the image block G _h is set to a predetermined value or less, that is, the predetermined number of image blocks G _h.
A processing unit consisting of _h has a fixed length, and each image block G _h
DCT of the image data and the obtained conversion coefficient C _ij is quantized and recorded, each image block G _h is regarded as a moving image block based on the conversion coefficient D _ij obtained by one-dimensional DCT in the vertical direction. The coding area is classified into still image blocks, the coding area is defined as a low frequency component area in the still image block, and the low frequency component and high frequency component area is defined in the moving image block, and only the transform coefficient C _ij of the coding area is quantized. , The classification information of each image block G _h , the quantization number m, and the area information indicating the coding area are recorded, and the inverse quantum based on the classification information, the quantization number m, and the area information at the time of reproduction. By performing the encoding and the reproduction, the encoding is efficiently performed at the time of recording, so that a good image quality can be obtained.
Further, since the processing unit is a fixed length, editing and variable speed reproduction can be easily performed.

[0053]

As is apparent from the above description, in the present invention, the image data is divided into blocks each having n × n in the spatial arrangement as one block, and the image data of each block is orthogonalized using a cosine function. Transform to calculate transform coefficient, quantize this transform coefficient to form quantized data, and output the quantized data, based on the transform coefficient, each block is moving video block or no moving block Motion is detected by detecting whether it is a still image block, quantizing the low frequency component of the transform coefficient when the block is a still image block, and quantizing the low frequency component and high frequency component of the transform coefficient when the block is a moving image block. A memory for detection, a circuit for detecting a difference, and the like are unnecessary, and the circuit scale can be reduced as compared with the conventional device. In addition, motion detection is performed in two-dimensional D
The processing time can be shortened by performing the processing in parallel with the CT processing. Furthermore, the motion detection can be performed with high accuracy, and the coding efficiency can be improved.

[Brief description of drawings]

FIG. 1 is a block diagram showing a circuit configuration of an image encoding device to which the present invention has been applied.

FIG. 2 is a block diagram showing a circuit configuration of a recording system of a digital video tape recorder to which the image encoding device is applied.

FIG. 3 is a block diagram showing a circuit configuration of a reproduction system of a digital video tape recorder to which the image encoding device is applied.

FIG. 4 is a diagram showing a specific configuration of a motion detection circuit that constitutes the image encoding device.

[Fig. 5] Fig. 5 is a diagram illustrating a region of a transform coefficient for explaining a coding region of a still image block.

[Fig. 6] Fig. 6 is a diagram illustrating a region of a transform coefficient for explaining a coding region of a moving image block.

[Fig. 7] Fig. 7 is a diagram illustrating a region of a transform coefficient for explaining the quantization width of the coding region.

FIG. 8 is a diagram showing a distribution of transform coefficients of a still image block.

FIG. 9 is a diagram showing a distribution of conversion coefficients of moving image blocks.

[Explanation of symbols]

12 ... blocking circuit 13 ... DCT circuit 14 ... quantization circuit 14b ... region determining unit 14c ... code amount calculating circuit 50 ... motion detection circuit Q _1, Q _2, Q ₃ ... Quantizer

Claims (1)

Claim: What is claimed is: 1. A block forming means for dividing image data into blocks each of which has n * n pieces in a spatial arrangement, and image data of each block from the block forming means is set to a cosine function. Discrete cosine transform means for orthogonally transforming to calculate transform coefficients, quantizing means for quantizing transform coefficients from the discrete cosine transform means to form quantized data, and outputting the quantized data, Based on the transform coefficient from the discrete cosine transform means, it is detected whether each block is a moving image block or a still image block that does not move, and when the block is a still image block, the low frequency component of the transform coefficient is quantized. When the block is a moving image block, it has a control means for controlling the quantizing means so as to quantize the low frequency component and the high frequency component of the transform coefficient. Image coding device.