JP2792470B2 - Motion detection circuit - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a motion detecting circuit, and more particularly to a motion detecting circuit for detecting a motion vector in the compression and expansion of a video signal.

[0002]

2. Description of the Related Art FIG. 3 is a block diagram showing an example of a conventional motion detection circuit. This motion detection circuit includes difference circuits ₁₁ to 1
₃ , a comparison circuit 2, a latch 3, and a selector 4. Differential circuit 1 ₁ to 1 horizontal direction 8 pixels ₃ each in the longitudinal direction 8 pixels (i.e., 8 × 8) data (reference data) of or of 8 × 16 reference block is a common input, and in the vertical direction N blocks Search data of each area when the area is divided into the search area of the M blocks in the horizontal direction is input, and difference data vector data between the reference data and the search data is generated, and the difference in the vertical and horizontal directions is generated. The minimum value of the data and the vector data at that time are output. Assuming that the difference data is a total of 4 bits having the most significant bit as the sign bit in each of the vertical and horizontal directions, the difference data is in the range of +7 to -8 (where + is the downward direction of the screen and the rightward direction). ).

[0003] Comparison circuit 2 compares the minimum value of the output difference data of the difference circuit 1 ₁ to 1 _3, determined their minimum,
Detecting the minimum difference data in the N × M search area,
Output as a select signal. Also, the difference circuits ₁₁ to 1
Output vector data ₃ is supplied through the latch 3 to the selector 4, wherein the difference data is minimal region by a select signal, i.e., the smallest vector data movement is selected and output.

[0004]

[SUMMARY OF THE INVENTION] However, in the conventional motion detecting circuit described above, when the spread of the search area, approximately difference data of the same value is output from all of the difference circuit 1 ₁ to 1 ₃ in such a flat pattern Is increased, the difference data due to noise included in the video, A / D conversion, and a slight error generated at the time of Y / C separation for separating the luminance signal and the carrier chrominance signal from the composite video signal. There is a drawback that the influence is increased, erroneous detection is increased, a large vector is output, and as a result, the information amount is increased.

[0005] The present invention has been made in view of the above points,
It is an object of the present invention to provide a motion detection circuit capable of reducing the frequency of erroneous detection and reducing the amount of information.

[0006]

According to the present invention, in order to achieve the above object, pixel data of different search ranges are input and data of a common reference block are input. A plurality of difference circuits for outputting the difference data and the vector data thereof, and output difference data of the plurality of difference circuits are provided as input signals and are respectively input.
The larger the difference data search range is to the center coordinate, the larger the value
A plurality of weighting circuits for performing weighting, a comparison circuit for detecting minimum difference data among output difference data of the plurality of weighting circuits, and a comparison circuit among vector data output from the plurality of difference circuits, respectively. And selecting means for selecting vector data corresponding to the minimum difference data.

[0007]

According to the present invention, the difference data between the pixel data in the search range and the data in the reference block is weighted according to the search range. By performing larger weighting on the difference data with the data of
The frequency of selection of the vector data of the difference data near the center position of the search range by the selection means can be increased.

[0008]

Next, embodiments of the present invention will be described with reference to the drawings. FIG. 1 is a block diagram of an embodiment of a motion detection circuit according to the present invention, and FIG. 2 is an explanatory diagram of the operation of FIG. As shown in FIG. 1, the present embodiment the differential circuit ₁ 1 to 1 _3, a latch 3,
Weighting circuits ₅ 1 to 5 _3, is configured more comparison circuit 6 and the selector 7, compared with the differential circuit ₁ 1 to 1 ₃ circuit 6
It is characterized in that provided weighting circuit 5 ₁ to 5 ₃ between. The weighting circuit ₅ 1 to 5 ₃ to weight of (1 + α).

Next, the operation of this embodiment will be described.
N blocks in the vertical direction and M blocks in the horizontal direction (N × M)
Of the search area, while the differential circuit 1 ₁ to 1 ₃ delimit an area in a range that can be searched, the pixel data of the moving image of the respective areas (search data) are input to the differential circuit 1 ₁ to 1 _3, vertical 8 pixels in the direction, 8 pixels in the horizontal direction (8 ×
8) or 8 pixels in the vertical direction and 16 pixels in the horizontal direction (8 × 1
Data of the reference block 6) (reference data) is inputted in common to the differential circuit ₁ 1 to 1 _3.

For example, as shown in FIG. 2, an 8 × 16 reference block 11 is located at the center of the search area, and search data blocks 12, 13 and 1 indicated by a dashed line are shown.
4 are at the positions shown in the figure. here,
Each of the search data blocks 12, 13 and 14 has three vertical and two horizontal blocks having the same size as the reference block 11, respectively.
And the search data included in the search data 14 is data one field (or one frame) earlier than the reference data.

A first included in the block 12, 13 and 14, second and third search data are input respectively separately to the differential circuit ₁ 1, 1 ₂ and 1 _3, references from the reference block 11, where The difference between the data and the vector data is obtained from the minimum difference data and the vector data obtained by this.
Weighting circuits 5 ₁ to 5 provided corresponding difference data
₃ and input the vector data to the latch 3 in common.

[0012] For example, in FIG. 2, block and vector data of the smallest difference data obtained arrow 16 and block 15 surrounded by a broken line by the difference circuit 1 _1, the arrow 18 block 17 enclosed by a broken line differential circuit 1 in block and vector data of the smallest difference data obtained by _2, the arrow 20 block 19 surrounded by a broken line is a block and vector data of the smallest difference data obtained by the differential circuit 1 _3.

[0013] weighting circuit ₅ 1 to 5 _3, and outputs the differential data (1 + alpha) multiplied by the comparator circuit 6. Here, for the output difference data of the difference circuit whose search range includes the center position of the xy coordinates (0, 0), the value of α is set to “0”.
The value of α increases as the output difference data of the difference circuit moves away from the center position in the search range.

[0014] Comparison circuit 6 is input weighting circuit _{5 1}
Receiving to ₅₃ of the output difference data as an input signal, detects the minimum difference data from these outputs a select signal. That is, as the output difference data of the comparator circuit 6 weighting circuit input is 5 ₁ to 5 _3, compares the sum with each other the values of all the pixels after each weighting in block 15, 17 and 19 in FIG. 2, they Is detected as the minimum difference data, and a select signal indicating the block is output.

On the other hand, the latch 3 temporarily stores the input vector data and supplies it to the selector 7.
The selector 7 selects and outputs vector data corresponding to the minimum difference data according to a select signal from the comparison circuit 6. Therefore, for example, assuming that the block 17 in FIG. 2 is detected as the minimum difference data by the comparison circuit 6, the selector 7 selects and outputs the vector data 18. The output vector data is transmitted after being converted into a variable length code (for example, Huffman code) according to its size.

With such a configuration, the difference data input to the comparison circuit 6 becomes larger than the actual value as the distance from the center position of the search range increases. Therefore, when the difference data outputted by each of the differential circuit 1 ₁ to 1 ₃ differ only slightly, the weighting circuit 5 ₁
The weighting operation by the 5 _3, the frequency of the vector of the difference data is close to the center position is output from the selector 7 is increased.

[0017] Indeed, in the flat pattern in theory should difference data of the same value from each of the differential circuit 1 ₁ to 1 ₃ are output, the differential data due to the effects of noise slightly different Is output, and the difference data farthest from the center in the search area is minimized, which may result in a large vector. However, according to this embodiment, the weighting operation by the weighting circuit 5 ₁ to 5 _3, since the vector of difference data close to the center position increases the frequency output from the selector 7, to prevent such erroneous detection be able to.

As a result, in contrast to the conventional case where a large amount of compressed code is transmitted according to a vector larger than the original value, in the present embodiment, a small amount of compressed code is output according to a small vector. Thus, the amount of information to be transmitted can be reduced.

The present invention is not limited to the above-described embodiment. For example, the difference circuit and the weighting circuit are not limited to three circuits.

[0020]

As described above, according to the present invention,
By weighting the difference data between the pixel data in the area farther from the center position of the search range and the data of the reference block more heavily, the frequency at which the vector data of the difference data closer to the center position of the search range is selected by the selection means is reduced. Since it is set high, even when the search range is expanded, the frequency of erroneous detection of motion detection caused by the influence of noise, A / D conversion, Y / C separation, and the like can be reduced, and the center position (0, 0) Can be generated more frequently. Thus, according to the present invention, the amount of information to be transmitted can be reduced in the compression and decompression of a video signal.

[Brief description of the drawings]

FIG. 1 is a block diagram of one embodiment of the present invention.

FIG. 2 is an operation explanatory view of a main part of FIG. 1;

FIG. 3 is a block diagram of an example of the related art.

[Explanation of symbols]

1 ₁ to 1 ₃ differential circuit 3 latches 5 ₁ to 5 ₃ weighting circuit 6 the comparator circuit 7 Selector

Claims (1)

(57) [Claims] 1. A plurality of difference circuits for receiving pixel data of different search ranges and data of a common reference block, taking a difference between these input data, and outputting minimum difference data and vector data thereof. , Provided corresponding to the plurality of difference circuits, receiving output difference data of the plurality of difference circuits as an input signal,
The farther the search range of the input difference data is from the center coordinates,
A plurality of weighting circuits for performing weighting of which value; a comparison circuit for detecting minimum difference data among the output difference data of the plurality of weighting circuits; and a vector data respectively output from the plurality of difference circuits. Selecting means for selecting vector data corresponding to the minimum difference data detected by the comparing circuit.