JPH0289187A - Picture signal correcting device - Google Patents

Abstract

PURPOSE:To eliminate the deformation of the shape of the boundary part of a picture by setting two threshold value, and outputting the binary data of the picture element of a sampling area as it is when the density o;f the sampling area is between these threshold value. CONSTITUTION:For every picture element of the binary picture signal stored in a memory 101, a sampling area setting means 102 sets the sampling area whose center is approximately said picture element, and a ratio calculating means 103 reads picture element data corresponding to the set sampling area from the memory, and counts the picture element data to show '1' and '0' contained in it, and calculates the ratio of the picture element data '1' in the area. When the ratio of the number of the picture element data to show '1' in the sampling area exceeds a set range, outputting means 104 to 106 make a corresponding center picture element into '1', and when the ratio is below the set range, they make the center picture element into '0', and when the ratio is within the set range, they output the data as it is. Thus, the shape of the boundary part of the figure can be kept correct even after correction.

Description

Detailed Description of the Invention (Industrial Field of Application) The present invention is directed to the use of a binary image signal obtained by scanning and imaging an object for surface inspection of an object such as an automobile body outer panel. The present invention relates to an image signal correction device for removing noise that is present in images.

(Prior art) Conventionally, as this type of image signal correction device, Japanese Patent Application Laid-Open No. 59-1
The one shown in Japanese Patent No. 78872 is known. As shown in FIGS. 7 to 9, this is done by setting a small area of mxn pixels centered around the pixel P to be corrected in the human input image signal. The image signal corresponding to this small area is read out, and the pixel data "l" contained therein is counted. As shown in FIG. 8, if the number is greater than or equal to the threshold Th, the pixel P is
l”, the pixel P is determined to be “0” and output. As a result, the image before correction shown in FIG. Noise figures indicated by speckles are deleted.

(Problems to be Solved by the Invention) However, although such a conventional image signal correction device removes noise well, as is clear from FIG. Therefore, when attempting to inspect the shape of an article based on manually generated image signals, this image signal correction device is used because the critical periphery of the image is deformed. The problem is that I can't.

(Object of the Invention) This invention was made to solve the problems of the conventional art, and its purpose is to create an image in which the shape of the boundary of the image is less likely to change even after correction. An object of the present invention is to provide a signal correction device.

(Means for Solving the Problems) In order to achieve the above object, the present invention is constructed as shown in FIG. 1.

The memory 101 stores a binarized image signal obtained by comparing and discriminating a video signal with a predetermined level when input thereto.

The sampling area setting means 102 sets, for each pixel of the binary image signal stored in the memory 102, a sampling area approximately centered on that pixel.

The ratio calculation means 103 reads out pixel data corresponding to the set sampling area from the memory, counts the pixel data indicating "lo" and "0" contained therein, and calculates the pixel data "1" in the area. Calculate the ratio.

If the ratio of the calculated pixel data "l" is within a preset range, the output means 104 reads out the pixel data corresponding to the center pixel from the memory 101 and outputs it as is, so that the pixel data corresponding to the center pixel Data.

If the ratio of the calculated pixel data "l" exceeds a preset range, the output means 105 outputs "l" as pixel data corresponding to the center pixel.

If the ratio of the calculated pixel data 11@ is less than a preset range, the output means 106 outputs "0" as pixel data corresponding to the center pixel.

(Function) In this invention, when the ratio of the number of pixel data indicating "1" in the sampling area exceeds the set range, the corresponding center pixel is set as "l", and when it is less than the set range, it is set as "0". By outputting manually entered data as is if it is within a set range, the shape of the boundary of a figure can be accurately maintained even after correction.

(Embodiment) FIG. 2 is a block diagram showing the configuration of an embodiment in which the image signal correction device according to the present invention is applied to a device for evaluating waviness and distortion of an object surface, that is, a panel such as an automobile body outer panel. .

In the figure, the striped pattern projected from projector l is
The light passes through the diffusion plate 2 and is irradiated onto the object 3 to be inspected. Tested object 3
The striped pattern projected onto the surface of is scanned and imaged by the imaging device 4, and outputted as an analog signal (2) corresponding to the brightness of the image.

This analog signal V is stored in the image memory A after being converted into a digital signal by the to/D converter 5.

The digital signal A stored in the image memory 8 is sent to the binarization processing unit 7, where it is compared and discriminated with a predetermined level in pixel units, and is converted into a binarized image signal B consisting of "1" or "O".
After being converted into , it is stored in image memory B.

The binarized image signal B stored in the image memory B is sent to the binary image correction unit 9, and is converted into a corrected image signal C by a correction process described below on a pixel basis, and then stored in the image memory C. .

The corrected image signal C stored in the image memory C is sent to the evaluation parameter calculation unit 11, where evaluation parameters for evaluating the smoothness of the object surface are calculated, and sent to the output unit 12, which includes a printer, floppy desk, CRT, etc. sent and output.

Note that the evaluation of the smoothness by the evaluation parameter calculation unit 11 is as follows:
This is done by extracting the edge line data of the slit image by the method described in Japanese Patent Application No. 1983-304 filed by the present applicant. That is, the distance 1[11Ds between the starting point PMT and the ending point PtH of the extracted edge line, and the distance D of the edge line itself.
The ratio DS/DL to L is calculated as a parameter, and the closer the value of this parameter is to 1, the smoother it is determined.

FIG. 3 is a diagram showing a processing target range of the binary image signal B processed by the binary image correction unit 9.

As shown in the figure, the binarized image signal B (x, y) expressed in x, y coordinates has a starting point (a, b) and an ending point (c, d).
Correction processing is performed on a correction area consisting of a rectangle determined by.

FIG. 4 is a flowchart showing the processing of the binary image correction section 9.

As shown in the figure, the correction process starts at the starting point of the correction area (
For each pixel from a, b) to the end point (c, d), a micro area (sampling area) of 21 pixels in the X direction and 5 pixels in the y direction, a total of 10,505 pixels, centered on that pixel.
This is done by counting the pixel data "1" contained therein.

The counted value is set to a high threshold (a65) as shown in FIG.
That is, if the total number of pixels exceeds 65 out of 10,505 pixels, the corrected image signal C is set to "lo".

When the count value is less than the low threshold 40, that is, 40 pixels out of a total of 10505 pixels, the corrected image signal C is set to "0". When the count value is the low dark value 40 or more and the high dark value 65 or less, the corrected image signal C is set to 2. (The corresponding pixel data of the a-ized image signal B is used as the corrected image signal C.

The darkness value set here is based on the state of image data obtained from the inspected object 3 and the evaluation parameter calculation unit 1 in the post-process.
The optimum value is set according to the request of No. 1.

FIG. 6 is an explanatory diagram showing an example of processing obtained in the example, in which figure a is a screen showing the binarized image signal B before correction;
In Figure 0, which is a screen showing the corrected image signal C after correction, the noise figure represented by minute spots that had occurred on the image is deleted, and at the same time, the shape of the periphery of the figure in the image is changed. It is shown that it remains the same without any change.

As described above, according to the correction device of the embodiment, noise removal processing can be performed in a short time as in the conventional method as a pre-processing for smoothing evaluation. In addition, the peripheral shape of the image at that time hardly changes, and the evaluation parameter calculation unit 11
The influence on smoothness evaluation is also eliminated, making it possible to obtain highly accurate evaluation results.

Note that in the binary image correction unit 9 of the embodiment, pixel data “1”
However, since this ratio is the complement of the ratio of pixel data "0", it goes without saying that any ratio has the same meaning.

(Effects of the Invention) The present invention sets two different values (a) when correcting to remove noise contained in a binary image signal as described above.
When the density of the sampling area is between these dark values, the binary data of the corresponding pixel is output as is, resulting in an image with almost positive values without changing the shape of the image boundary. The effect of being able to obtain a signal is obtained.

[Brief explanation of drawings]

FIG. 1 is a complaint correspondence diagram, FIG. 2 is a block diagram showing the configuration of an embodiment according to the present invention, FIG. 3 is a diagram showing the processing target range of the binary image signal B, and FIG. 4 is a binary image Correction section 9
A flowchart showing the processing of
FIG. 6 is an explanatory diagram showing the change in the image before and after processing, and FIG. 7.8.9 is a diagram showing an example of conventional processing. 101...Memory 102...Sampling area setting means 103...
... Ratio calculation means 104 ... Output means lO5 ... Output means 106 ... Output means! ...Projector 2...Diffusion plate 3...Object to be inspected 4...Imaging device 5...A/D converter 6...Image memory A 7...2 Value conversion processing section 8... Image memory B 9... Binary image correction section IO... Image memory C 11... Evaluation parameter calculation section 12... Output section patent patent Counsel: Nissan Motor Co., Ltd. Patent Attorney Shigenori Wada

Claims (1)

[Claims] 1. A memory that stores the binarized image signal obtained by comparing and discriminating the video signal with a predetermined level as pixel data, and a sampling area approximately centered on that pixel for each pixel of the pixel data stored in this memory. A setting means for setting and reading pixel data corresponding to the set sampling area from the memory and
a calculation means that counts pixel data indicating ``1'' and calculates the ratio of pixel data ``1'' within the area; An output means that outputs the content of pixel data corresponding to a pixel as it is as pixel data corresponding to a center pixel, and if the ratio of calculated pixel data "1" exceeds a preset range, it is output as pixel data corresponding to a center pixel. If the ratio between the output means that outputs "1" as pixel data and the calculated pixel data "1" is less than a preset range,
An image signal correction device comprising: output means for outputting "0" as pixel data corresponding to a center pixel.