JPS60135707A - Detecting method of ruggedness defect - Google Patents

Abstract

PURPOSE:To detect even a ruggedness defect which can not be detected by binary coding processing by comparing the difference of the mean value of brightness of image data in each partitioned area with a reference value. CONSTITUTION:The surface of a body to be inspected is irradiated with regularly reflected light from a light source and a regularly reflected image of the body to be inspected is shot by a television camera 3. The output of the television camera 3 is A/D-converted by an A/D converter 6 and then stored in a frame memory 7. A computer 8 sections image data stored in the frame memory 7 horizontally and vertically, calculates the mean value of the brightness of image data in respective sectioned area, and comparing the difference of the calculated difference with the reference value to detect the ruggedness defect of the objective body.

Description

[Detailed Description of the Invention] [Technical Field] The present invention captures uneven defects on the surface of an object to be inspected into an image using a television camera, and performs multivalue numerical processing using a This is a method 131J for detecting irregularities on the buttocks.

[Background technology]

In the conventional unevenness defect detection method, first, a specularly reflected image obtained by applying specularly reflected light to an object to be inspected is processed by a binarization process such as floating binarization. Here, floating binarization involves comparing input signals at a certain threshold value and converting them into binary values (for example,
, O), the threshold value changes according to the number of input cards. The images were then put into a computer and defects were detected using pixel counting and matching techniques. That is, after the image is binarized, a mask is prepared for the collection of images, and the original image is determined using this mask.

In other words, defects were detected by overlapping a mask on the original image (matching) and counting the number of pixels in the area where the mask and the original layer overlapped (pixel count). However, unevenness defects on the surface of the object to be inspected occur at random locations, and the matching method cannot be used. That is, the Matsuchinji method cannot be used unless the object to which the method is applied is known, or in other words, the mask cannot be defined. Furthermore, since the uneven defects are not uniform in size, the method of counting the number of pixels is also inaccurate.

[Purpose of the invention]

An object of the present invention is to enable processing using the same criteria even if the size of defects is not uniform, and to detect uneven defects that are difficult to detect with conventional binarization processing. It's about doing.

[Disclosure of the Invention] Embodiment In FIG. 1, il) is a light source that illuminates the object to be inspected (2) obliquely from above and provides specularly reflected light to the surface of the object to be inspected (2). (3) is a television camera that captures a regular reflection image of the object to be inspected (2). At this time, the regular reflection image of the convex defect becomes a bright part (4) on the A side and a dark part (5) on the B side, as shown in FIG. As shown in FIG. 3, the video signal from the television camera (3) is input to an A/D converter (6), A/D converted, and sequentially written into a frame memory (7). (8) is a computer that numerically processes the image data of the object to be inspected (2) written in the frame (7). (9) is a 0-frame meme that writes processing results and intermediate results (January 7) is a TV camera (3
) is a high-speed, large-capacity memory that can write image data for one image in synchronization with 256 x 2
! The memory is distributed over 56 pixels, and one pixel is 256 pixels.
It has light amount data of gradations (8 pits).

The 256 x 256 pixels as shown in Fig. 4(a) written in the operation frame message (7) are divided into N in the horizontal direction and 1 in the vertical direction as shown in Fig. 4(b). Divide into N x M areas. In the case of Fig. 4(b), it is divided into 64 regions with M=8 and N=8, and the number of pixels in one region is 32.
X'32 = I 024 pixels. The average amount of light for the 1024 pixels in this one area is calculated by a computer (8) and written in one page. This is 6
Calculations are made for all four areas, and the calculated average light quantity is assigned as Kn and KI2 along the position of the area as shown in FIG. From now on, the position will be indicated by the average light intensity K M'm.

Next, for the average light intensity of one @ area, 8
Nearby, for example, 42, K43, K44 for K53
, K52, K54, K112, K63, K64 8
It compares it with the average light amount of the nearby trace, calculates the difference in light amount from KMN among the eight nearby light amounts that are smaller than the light amount of KMN, and selects the largest difference. This is called SMN. Expressed in the formula,
SMN=KMN-MIN (8 neighbors of KMN) However,
KMN −MIN (8 neighborhood of KIN)〉0, and MI
For N (8 neighbors of KMN), select the minimum value among the 8 neighbors.

This SMN also ranges from So to SBs. However, S.

Calculated from 3 neighbors of 21N K22 and KI2, 5
12 is calculated from 5 neighbors. SN+, SNg, 51M5
58M is calculated in the same way. If MIN (8 neighbors of KMII) does not satisfy the condition, SMN=0. Find the maximum value among the 64 SMNs calculated in this way.
The presence or absence of an uneven defect can be determined by comparing it with the criteria.

Specifically, as shown in FIG. 6(a), if there is a convex defect extending over 32 K22 and K22, there are 22 convex defects in the bright area.
The average light quantity of the dark area K is 100, and the average light quantity of the dark area K is 2.
Set to 0. Also, assuming that the remaining average light quantities are all 50,
As shown in FIG. 6(b), when s, , = go and S3λ = 0, a convex defect can be detected.

[Effects of EJJ originating]

As described above, the present invention includes a light source that provides specularly reflected light to the surface of an object to be inspected, a television camera that captures a specularly reflected image of the object to be inspected, and an A/D converter that converts the output of the television camera.
A/D converter, and a frame memory for sequentially storing the output of the A/D converter, and divides the image data stored in the frame memory into N pieces in the horizontal direction and M pieces in the vertical direction. means for dividing the image into M x N regions; means for reading the image data of the one region and calculating the average value of brightness of each point; and means for comparing the average values of each region and calculating the difference; Since uneven defects are detected using a comparison means that compares the difference between the calculated average values with a reference value, even if the defects are not uniform in size, they can be processed using the same criteria, making it difficult to detect with conventional binarization processing. This has the effect that rough unevenness defects can be detected.

[Brief explanation of drawings]

FIG. 1 is a perspective view of an embodiment of the apparatus used in the detection method of the present invention, FIG. 2 is a front view of an example of a specularly reflected image of the above television camera, and FIG. 3 is a block circuit diagram of the same above. vJ4 diagram (a
L (b) is an explanatory diagram of the image data written in the frame memory, Figure 5 is a diagram showing the average light intensity allocation of the same as above, and Figure 6 (
a) and (b) are Maki explanatory diagrams of the unevenness defect detection method of the present invention. (II...Light source, (2)...Test object, (3)...
・Teshiji camera, (6)...A/D converter, +7+・
...framework, (8) ...computer, (9
)···memory. Agent Patent Attorney Ishi 1) Chief Figure 1 and Figure 2 Figure 3 Figure 4v4 (0) (b) No. 514 6th cause (o) (b) ゛

Claims (1)

[Claims] (1) A light source that provides specularly reflected light on the surface of the object to be inspected, and
A television camera that captures a specularly reflected image of a fertilizer liquid test object; 1 an A/D converter that converts the output of the television camera from A/D to digital; and a frame memory that sequentially stores the output of the A/D converter. The image data stored in the frame is divided into N pieces in the horizontal direction and M pieces in the vertical direction, dividing one image into MXN areas (one means and an image arc of the one area). A means for calculating the average value of the brightness of each readout point, a means for calculating the difference by comparing the average values of each area, and a comparison for comparing the difference between the calculated 1,000 average values with a reference value. By means of [uneven defect detection/
1 Out, p method.