JPS6238347A - Image processor - Google Patents

Abstract

PURPOSE:To shorten an image processing time and to reduce the capacity of an image storage memory by converting image information from analog to digital into a binary signal and compressing binary-coded information of every picture properly. CONSTITUTION:A two-dimensional scanner 3 is moved by an XY two-axis robot 4 in a device which measure an image pattern on paper, etc., so as to cover the entire area on a body to be measured which is paper, etc., (obtained by photographing a flaw made in, for example, a manufacture process of metal). Then, the video signal output of the scanner 3 is converted by an A/D converting circuit 5 into a multi-valued signal, which is then converted into a binary signal by a binary-coding circuit 6 which employs a mode method, etc. Further, the binary-coded information is compressed by a compressing circuit 7. This compressed information is sent to an image processing circuit 8, which fetches information of every visual field and then sends a next movement instruction to the robot 4. This operation is performed repeatedly as many time as the number of times of compression and then an image processing circuit 8 outputs the number and area of flaws to the outside.

Description

DETAILED DESCRIPTION OF THE INVENTION [Field of Industrial Application] The present invention relates to an image processing device that measures image patterns on paper sheets.

[Conventional technology]

Conventionally, the area or number of individual fragments has been counted manually from the image pattern on paper II.

[Problem that the invention seeks to solve]

]-The conventional human measurements described above have the drawbacks that measurement errors occur depending on the level of inexperience or fatigue of the individual, and that it takes time to measure the area. Various image processing devices have been proposed to eliminate this drawback. However, in a typical image processing device, the measurement area is fixed, and the higher the processing speed per screen, the higher the price. Therefore, when measuring a wide area, either adjust the distance between the scanner and the object to be measured, or move the object or scanner while keeping the distance between the scanner and the object fixed (all measurement areas Processing was performed at a frequency of ÷measurement area per screen). In this case as well, the former has the disadvantage that the measurement accuracy is reduced due to the low resolution, while the latter takes time, and movement errors cause small shifts in the measurement area for each screen. This has the disadvantage that it takes more time for correction.

[Means for solving problems]

The image processing device of the present invention includes a two-dimensional scanning type scanner that scans one paper sheet two-dimensionally, a moving means that moves the scanner in the two-dimensional direction of the paper sheet, and an A/D converter that converts the output of the scanner into an A/D converter.
means for converting, means for determining a threshold value for binarization from the original image information obtained by A/D conversion, and a means for determining a threshold value for binarization based on the original image information obtained by A/D conversion, and means for compressing the binarized information.

In this way, by A/D converting the image information, binarizing it, and appropriately compressing it, the image processing time is shortened and the image storage memory capacity is slightly reduced.

〔Example〕

Next, embodiments of the present invention will be described with reference to the drawings.

FIG. 1 is a block diagram of an embodiment of an image processing apparatus of the present invention, and FIG. 2 is a diagram showing an object to be measured l.

Assume that the object to be measured l is, for example, a photograph of a flaw that occurs in a metal manufacturing process. Note that this example describes scratches on metal surfaces as a specific example, and it is natural that it can also be applied to other applications, such as the extraction measurement of cell boundaries. has a size of XPIAx
The size of the two-dimensional scanning scanner 3, which will be described later, has a size of X YM^I.
Assume that there is only the size of Yscan. Here, there is a flaw 2a on the object to be measured l.

There are 2b, 2c, 2d, 2e, and 2f, and in order to calculate the total number and area of scratches, use 2 as shown in Figure 2.
A dimensional scanning type scanner 3 is moved by an XY two-axis robot 4 so as to cover the entire area (XM^!IIYMA) of the object to be measured.
YMA! = 100mm, X5can=
long, Yscan = 10m■ binding,
Assuming that the two-dimensional scanning type scanner 3 has a resolution of 1024 x 1024, the amount of image storage information required is XNA x m Ynax if repeated measurements are made as is.
/ X5cana Yscan X l024X l
024= Not only is the size greater than 108 megabytes, but the time required for image processing is also extremely long. Therefore, the video signal output of the two-dimensional scanning type scanner 3 is multivalued by the A/D conversion circuit 5, and then binarized by the binarization circuit 6 using the mode method or the like.
The digitized information is compressed by a compression circuit 7.
The compression ratio is X5can/XMA! is the X direction Ysc
Let it be an/YsAx. however.

This compression ratio may be changed depending on the amount of information that can be stored in the image, etc. It is also desirable to change the compression as follows depending on the object to be measured. If the part without scratches (ground color) is white and the scratched part black, the scratched part will be smaller compared to the total area. Therefore, in this case, priority is given to black. Note that in the opposite case to this example, priority is given to white. Giving priority to black means that the compression ratio in this embodiment described above is 1/10 in the X direction.
, Y direction 1/10 (7) II converts the original binarized information to X direction 1
This indicates that a method is used in which the compression result is black when there is black above a certain level in 0 pixel and 10 pixels in the Y direction. By performing this compression, A/D conversion or 2 (
In addition to compensating for lines that are blurred due to the inability to correct density differences during indium conversion, it is also possible to compensate for deviations in the field of view for each screen due to movement errors of the robot 4. This compressed information is sent to the image processing circuit 8, and after the image processing circuit 8 has taken in the information for each field of view, it issues the next movement command to the XYZ axis robot robot 4. After performing this operation for the number of compressed surfaces, the image processing circuit 8 outputs the number and area of scratches to the outside. Note that if you need area information of a certain size for such scratches, you can leave it as is, but if you want to increase the accuracy, you can use a two-dimensional scanning type based on compressed information. scanner 3
It is sufficient to move the image to the center of the flaw and process it without going through the compression circuit 7.

Effects of IJJ from C] As explained above, the present invention reduces the processing time for one image and reduces the image storage memory by A/D converting image information, binarizing it, and appropriately compressing it. effective.

[Brief explanation of drawings]

FIG. 1 shows the configuration IA of an embodiment of the image processing device of the present invention.
214 is a plan view of the object to be measured l. l...Object to be measured, 2a, 2b, 2c, 2d, 2e, 2f - Scratch, 3...
・Two-dimensional scanning scanner. 4...XYZ axis robo robot...A/D conversion circuit, 6...Binarization circuit, 7...Compression circuit. 8... Image processing circuit.

Claims (1)

[Claims] A device for measuring image patterns on paper sheets, comprising: a two-dimensional scanning scanner that scans the paper sheets two-dimensionally; and a moving means that moves the scanner in the two-dimensional direction of the paper sheets; means for A/D converting the output of the scanner;
A means for determining a threshold value for binarization from the original image information obtained by /D conversion, and a means for compressing the binarized information according to the area ratio between the maximum measurement area and the viewing area of a two-dimensional scanning type scanner. An image processing device comprising: