JP2712826B2 - Motion detection device - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION [Industrial applications]

The present invention relates to a motion detection device that detects a motion of an image in a video printer or the like that prints an image based on a video signal, and particularly to an improvement in an error detection unit that detects an error in a determination result.

[Prior art]

Techniques for detecting and removing errors in image motion determination include:
Area filters are known and are often used in EDTV. This area filter is a kind of majority circuit, and detects an error of a motion determination result in a pixel of interest using a determination result of a peripheral pixel. For example, assume an area of m dots (corresponding to the horizontal direction of the television screen) × n lines (corresponding to the vertical direction of the television screen) as shown in FIG. The center is the target pixel 10 for error detection. First, the motion of each of the pixels included in this area is determined. Next, "movement"
The number of pixels determined to be is counted. This count value is compared with a set threshold value Th in a characteristic graph as shown in FIG. 6, for example. As a result, when the count value is equal to or smaller than the set threshold, the target pixel 10 is determined to be “still”, and when the count value is equal to or larger than the set threshold, the target pixel is determined to be “moving” (for example, the Television Society of Japan). See Technical Report, Vol. 12, No. 51, PP. This determination result is compared with the first determination result, and an erroneous determination is detected. That is, when the two do not match, it is determined as “error”.

[Problems to be solved by the invention]

However, in the above-described prior art, since simple majority is used as a basis for error detection, there is an inconvenience that erroneous detection or omission of detection occurs at a boundary portion between a moving portion and a still portion of an image. For example, in the area mentioned above,
It contains 25 pixels. Where 12 of them
It is assumed that each is a moving part and 12 are stationary parts. It is also assumed that the detection threshold value Th is 13. Then, whether the determination result of the target pixel is erroneous depends on whether the remaining one is moving or still. The present invention has been made in view of the above, and it is possible to satisfactorily detect an error of a motion determination result without malfunctioning even at a boundary portion between a moving portion and a still portion of an image, and particularly to improve image quality in frame printing. It is an object of the present invention to provide a motion detection device suitable for the above.

[Means for Solving the Problems]

The present invention provides a motion detection device that detects an error in a motion determination result of a target pixel using a motion determination result of a pixel included in a pixel region including the target pixel. For the target pixel, areas of adjacent pixels and peripheral pixels are provided, and when the determination results in the adjacent pixels are not continuously equal or when the determination results of the adjacent pixels and the peripheral pixels are not continuously equal, An error detecting means for making an error in the determination result of the target pixel is provided.

[Action]

According to the present invention, an error in a motion determination result is detected and corrected satisfactorily without malfunction even at a boundary portion between a moving portion and a still portion of a pixel. improves.

【Example】

Hereinafter, an embodiment of a motion detection device according to the present invention will be described with reference to the accompanying drawings. Note that the same reference numerals are used for portions that are the same as or correspond to the above-described conventional example. In the figure, a mark “○” indicates that the motion determination result of the pixel is “still”, and a mark “x” indicates that the motion determination result of the pixel is “movement”. First, a basic error detection method according to the present embodiment will be described with reference to FIG. One is that the determination of sporadic "movement" or "stillness" is clearly incorrect. Specifically, the determination result of the target pixel 10 is “movement” and the determination result of the peripheral pixel is “still”, and the determination result of the target pixel 10 is “still” and the determination result of the peripheral pixel is “movement”. , The determination result of the target pixel 10 is determined to be erroneous. In the example of FIG. 9A, the determination result of the target pixel at the center is “still”, whereas the determination result of the peripheral pixels is “movement”. The second is that the determination result of the target pixel is correct when a pixel having the same determination result continues in any of the upper, lower, left, right, and oblique directions of the target pixel 10. In the case where a telop flows on the television screen, pixels having the same “movement” determination result as the pixel of interest 10 are continuous, for example, as shown in FIG. In this embodiment, error detection is performed by such a method,
Further, the determination result of the target pixel 10 is corrected. Next, a configuration of a main part of the present embodiment will be described with reference to FIG. In the figure, the motion determination result of the target pixel 10 is supplied to the terminal 12 from outside. The determination results of eight pixels adjacent to the target pixel 10 (referred to as “adjacent pixels”) are supplied to terminals 14A to 14H, respectively. Terminals 14A to 14H are two-input EXNO
R gates 16A to 16H are respectively connected to one input side. Terminal 12 is connected to the other input side of EXNOR gates 16A to 16H.
Are connected to each other. The output side of these EXNOR gates 16A to 16H is connected to an 8-input OR gate 18, an 8-input encoder 20, 8
Each is connected to the input side of an input gate 22. Next, the output side of the encoder 20 is connected to the data selector unit 24.
Is connected to the input side. This data selector 24
Is composed of eight input data selectors 24A, 24B, 24C. These data selectors 24A to 24C have 16 pixels ("peripheral pixels") adjacent to the outside of the adjacent pixels.
) Is input. That is, the motion determination results of 16 pixels around the 5 pixel × 5 line area centered on the target pixel 10 are appropriately distributed and input to the data selectors 24A to 24C. In addition, three bits on the output side of the encoder 20 are connected to the select signal input sides of the data selectors 24A to 24C, respectively. The output side of the data selectors 24A to 24C is a 2-input EXNOR
Each of the gates 26A to 26C is connected to an input side. Terminals 12 are connected to the other input sides of these EXNOR gates 26A to 26C, respectively. The output side of EXNOR gates 26A to 26C
It is connected to the input side of a three-input OR gate 28, respectively. The output side of the OR gate 28 is connected to the OR gate 18 described above.
Are connected to the input side of a two-input AND gate 30, respectively. Next, the output side of the AND gate 30 is connected to the output side of the two-input NOR gate 32 together with the output side of the gate 22 described above. The output side of the NOR gate 32 is connected to the input side of the two-input EXOR gate 34 together with the terminal 12. The output side of the EXOR gate 34 is connected to the output terminal 36. In each of the above units, the EXNOR gates 16A to 16H and 26A to 26C perform a NOT operation of the exclusive OR, and output the logical value “H” when the input logical values match. Has become. In the encoder 20, based on the inputs of the EXNOR gates 16A to 16H, the position of the pixel whose logical value is "H" is 3
It is output in bits. That is, the position of the pixel of the same motion determination result as the target pixel among the eight pixels adjacent to the target pixel 10 with respect to the target pixel 10 is output to the data selector unit 24. I have. The gate 22 outputs "H" when the input has two or more logical values "H". The data selector unit 24 outputs, from the data selectors 24A to 24C, the motion determination results of three neighboring pixels adjacent to the pixel having the logical value “H” specified by the encoder 20. Next, the operation of the embodiment configured as described above will be described with reference to FIGS. Note that FIG.
Shows a specific example of the determination result, and FIG. 4 shows a flowchart of the main operation. By means not shown in the figure, 5 with the target pixel 10 at the center
The motion of the pixels included in the pixel × 5 line area is determined (step S10 in FIG. 4). Of these, the central pixel 10
Is supplied to the terminal 12. Also, the central pixel 10
, That is, the pixels (excluding the center pixel 10) within the frame shown by the thick line 40 in FIG. 3 are supplied to the terminals 14A to 14H, respectively. Further, the determination result of the peripheral pixels outside the frame indicated by the thick line 40 is supplied to the data selector 24. a, When there is no identical determination result in the adjacent pixels (Steps S12 and S18) First, when there is no identical result in the adjacent pixels with the determination result of the target pixel 10, ie, in FIG. A case as shown will be described. At this time, since the logical value of the terminal 12 does not match the logical value of the terminals 14A to 14H, the outputs of the EXNOR gates 16A to 16H have the logical value “L”. Therefore,
The outputs of the OR gate 18 and the gate 22 also become "L". For this reason, the output of the AND gate 30 becomes "L", the input of the NOR gate 32 becomes "L", and the output thereof becomes the logical value "H". Therefore, in the EXOR gate 34, the terminal 12
Is inverted and output. For example, as shown in FIG. 2, when the determination result of the target pixel 10 is “still”, it is inverted by the EXOR gate 34 and “movement” is output from the terminal 36. That is, the determination result of the target pixel 10 is determined to be erroneous, and is corrected and output. b, When there are two or more pixels having the same determination result in the adjacent pixels (steps S14 and S20) Next, in the pixels adjacent to the pixel of interest 10, there are two or more pixels having the same determination result as the pixel of interest 10 A case will be described. For example, the case shown in FIG. As shown in the drawing, adjacent pixels 42 and 44 having the same determination result as the target pixel 10 exist. At this time, EXNOR gates 16A-1
Two or more of the outputs of 6H have the logical value “H”. Therefore,
The outputs of the OR gate 18 and the gate 22 also become "H". Therefore, the output of the NOR gate 32 has the logical value “L” regardless of the logical value of the output of the AND gate 30. Therefore,
In the EXOR gate 34, the input from the terminal 12 is output as it is. That is, the determination result of the target pixel 10 is determined to be correct, and is output as it is. c, When one of the adjacent pixels has the same determination result and is not in the vicinity (steps S16 and S18). Next, among the pixels adjacent to the target pixel 10, those having the same determination result as the target pixel 10 There will be described a case where there is only one and there is no result of the same determination around the pixel. For example, the case shown in FIG. As shown in the figure, the pixel having the same determination result as the target pixel 10 has only the adjacent pixel 46. Peripheral pixel 4 adjacent to this adjacent pixel 46
8, 50 and 52 all have different judgment results. At this time, the output of the OR gate 18 becomes a logical "H" and the output of the gate 22 becomes a logical "L". Further, the encoder 20 outputs address information indicating the position of the adjacent pixel 46 to the data selector 24. In the data selector unit 24, based on the input address information, the determination results of the adjacent peripheral pixels 48, 50, and 52 are respectively determined by EXOR gates 26A, 26B, and 26C.
Is output to The EXNOR gates 26A, 26B, and 26C compare these determination results with the determination result of the central pixel 10 supplied from the terminal 12. In this case, since none of the determination results match, the outputs of these become the logical value “L”. Therefore, the output of the OR gate 28 becomes "L" of a logical value, and the output of the AND gate 30 also becomes "L". Therefore, the input of each of the NOR gates 32 becomes a logical "L", and the output thereof becomes a logical "H". Therefore, in EXOR gate 34, terminal 12
Is inverted and output. For example, when the determination result of the target pixel 10 is “movement” as shown in FIG. 3B, it is inverted by the EXOR gate 34 and “still” is output from the terminal 36. That is, the pixel of interest
The judgment result of 10 is judged to be incorrect, and is corrected and output. d, In the case where there is one identical determination result in the adjacent pixels and there are also neighboring pixels (steps S16 and S20). Next, among the adjacent pixels of the attention pixel 10, those having the same determination result as the attention pixel 10 Is described, and there is a case where the same determination result exists around the pixel. For example, the case shown in FIG. As shown in the figure, a pixel having the same determination result as the target pixel 10
45. Peripheral pixels 56, 58, 60 adjacent to this adjacent pixel 54
56 have the same determination result as that of the target pixel 10. At this time, address information indicating the position of the adjacent pixel 54 is output from the encoder 20 to the data selector 24.
In the data selector unit 24, based on the input address information, the determination results of the adjacent peripheral pixels 56, 58, and 60 are each EX.
Output to NOR gates 26A, 26B, 26C. In this case, since the determination result of the peripheral pixel 56 matches the target pixel 10, one of those outputs becomes the logical value “H”. Therefore, the output of the OR gate 28 becomes "H" of a logical value, and the output of the AND gate 30 also becomes "H". Therefore, one of the inputs of the NOR gate 32 becomes a logical "H", and the output thereof becomes a logical "L". Therefore, in the EXOR gate 34, the input from the terminal 12 is output as it is. That is,
The determination result of the target pixel 10 is determined to be correct, and is output as it is. As described above, according to the present embodiment, the determination of the sporadic “movement” or “stillness” for the pixel of interest is corrected as an error, and the same determination is made for either the top, bottom, left, right, or diagonal of the pixel of interest. The determination result of the target pixel when the result pixel continues is determined to be correct, and the correction is not performed. For this reason, even at the boundary between the moving part and the stationary part of the image, the filtering is performed satisfactorily, the singular point or noise is removed, and the image quality at the boundary is improved. If such motion detection is applied to frame printing using images of odd and even fields of a video printer, good motion correction is possible, and the image quality is improved. The present invention is not limited to the above embodiment. For example, in the above embodiment, a region of 5 × 5 pixels centering on the pixel of interest is targeted, but pixels of a further region may be processed by adding circuit elements. The target area is defined as m pixels × n lines (m ≠ n).
It may be. In addition, the circuit configuration can be variously changed in design such as using a ROM so as to achieve the same operation, and these are also included in the present invention.

【The invention's effect】

As described above, according to the present invention, adjacent pixel and peripheral pixel areas are provided for the target pixel at the boundary between the moving part and the stationary part of the pixel, and the determination result in the adjacent pixel is continuous. When they are not equal to each other, or when the determination results of the adjacent pixel and the peripheral pixel are not continuously equal, the determination result of the target pixel is determined to be an error. There is an effect that the error of the motion determination result is detected and corrected satisfactorily without malfunction, and the image quality of the frame print can be improved.

[Brief description of the drawings]

FIG. 1 is a block diagram showing an embodiment of a motion detection apparatus according to the present invention, FIG. 2 is an explanatory diagram showing a basic error detection method of the present embodiment, and FIG. 3 is a target pixel region and motion determination in the present embodiment. FIG. 4 is an explanatory diagram showing an example of the result, FIG. 4 is a flowchart showing the operation of this embodiment, FIG. 5 is an explanatory diagram showing an area filter,
FIG. 6 is an explanatory diagram showing the operation of the area filter. 10: Pixel of interest, 12, 14, 36 ... Terminal, 16A-16H, 26A-26
C: EXNOR gate, 18, 28: OR gate, 20: Encoder, 22: Gate, 24: Data selector section, 30:
AND gate, 32 …… NOR gate, 34 …… EXOR gate, 42,4
4, 46, 54 ... adjacent pixels, 48, 50, 52, 56, 58, 60 ... peripheral pixels.

Claims (1)

(57) [Claims] 1. A motion detecting apparatus for detecting an error in a motion determination result of a pixel of interest using a motion determination result of a pixel included in a pixel area including the pixel of interest. For the pixel of interest, the area of the adjacent pixel and the peripheral pixel are respectively provided, and when the determination results in the adjacent pixels are not continuously equal, or when the determination results of the adjacent pixel and the peripheral pixel are not continuously equal, A motion detection device comprising: an error detection unit that makes an error in the determination result of the target pixel.