JPH05290190A - Mark inspecting method - Google Patents

Abstract

PURPOSE:To properly judge a defect in an object image to be inspected even unless a binary-coding process is properly done by converting an input image signal into a digital signal and performing the binary-coding process, then extracting feature quantities of independent patterns which do not connect with each other in a reference image, and comparing the feature quantities of corresponding patterns of the reference image and image to be inspected. CONSTITUTION:A master image of a reference mark is inputted by using an input device such as a TV camera. Then this image is converted into binary data and the feature quantities of plural independent patterns in the image are calculated and registered together with their positions. When the object image is inspected, an image of a mark to be inspected is inputted by the TV camera and converted into binary data, and the feature quantities of independent patterns are calculated. The feature quantities are compared with the feature quantities of the independent patterns of the master image which are registered in advance and when the both are different, it is judged that the inspected object mark has a defect.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a mark inspection method, and is suitable for use in the field of detecting and recognizing differences and changes in target images based on reference image information.

[0002]

2. Description of the Related Art As a method for mechanically inspecting an object to be inspected, whether a mark formed by printing or etching is correct, an image input by a television camera or the like is used.
There are methods such as binarizing and using pattern matching between images, or comparing the feature amounts of a plurality of independent patterns in the images corresponding to each other between the reference image and the image to be inspected.

[0003]

When using the above-described method of comparing the feature quantities of a plurality of independent patterns,
It is necessary to binarize the input image. However, in the process of binarization, it may not be possible to properly binarize due to illumination unevenness, incompleteness of the binarization method, or unevenness in the density of the imaged object. At that time, when comparing the feature amounts of the corresponding independent patterns, a significant difference occurs,
The judgment (erroneous judgment) is made that there is a defect in the input image to be inspected.

In view of the above-mentioned problems, the present invention is to make it possible to appropriately judge whether or not there is a defect in the image to be inspected, even if the binarization processing to the image to be inspected cannot be performed properly. To aim.

[0005]

According to the mark inspection method of the present invention, the correspondence relationship between a plurality of corresponding independent patterns between the reference image and the inspection image is checked in advance, and the feature amounts of the corresponding patterns are compared. When a pattern is markedly different from each other, a feature amount of another pattern existing in the vicinity of the pattern is considered and compared with the pattern.

[0006]

[Function] Not only the corresponding independent patterns are compared but also a wider range of feature amount comparison including patterns existing in the vicinity is performed. Is not done,
For example, even if a character has a defect, the amount of the defect can be accurately evaluated. In addition, the above process is not performed for all independent patterns, but is performed only for the independent patterns generated between the reference image and the difference equal to or more than the preset allowable value. Almost no reduction in processing speed occurs. [Example] (Outline of inspection) FIG. 1 is a flowchart showing the procedure of the mark inspection method of the present invention. In the mark inspection method of the present invention, the reference mark is memorized once (registration of the reference image),
After that, it is a method of inspecting whether there is any defect by comparing with the inspection target mark.

The procedure of the mark inspection method is roughly divided into
It is composed of "registration of master image" and "inspection of target image". First, an image of a reference mark is input using an image input device such as a TV camera. Next, this image is binarized, the characteristic amounts of a plurality of independent patterns (particles) in the image are calculated, and the characteristic amounts and positions of each are registered. An example of the feature amount is shown in FIG. It is desirable that the feature quantity used here is one that is not affected by misalignment or rotation.

After the registration of the master image, the inspection target image is inspected. The image of the mark to be inspected is input by a TV camera or the like, binarized, and the feature amount of particles is calculated.

Then, as shown in FIG. 3, the feature amounts registered in advance when the master image is registered are compared for each particle. As a result of the comparison, if the feature amount differs from the master image, it is determined that the mark has a defect. After that, the inspection of the target image is similarly continued.

(Characteristic Matching) Next, the feature matching will be described in detail. In the collation of the feature amount, the feature amounts of the corresponding particles in the master image and the image to be inspected are compared with each other as shown in FIG. The feature quantity used here is as shown in FIG. 2, and in the present invention, the perimeter and the area are mainly used.

However, when the input image has density unevenness due to uneven lighting, or the threshold level at the time of binarization is not appropriate, the number of particles that was originally one becomes two or more. FIG. 4 shows an example in which the original character is separated into two due to poor binarization. In the case of the same figure, the character “N” in the master image on the left side is compared with the character “N” in the image to be inspected on the right side ”
The left side of N "is recognized as a corresponding particle. When the feature amounts are compared, a significant difference occurs even if the defective portion is sufficiently shorter than the allowable amount, and it is erroneously determined that the mark has a defect. To be done.

(Method of Recalculation When Cut Out) In order to avoid the above-mentioned inconvenience, the present invention recalculates the feature quantity by the following method and compares the feature quantities. When there is a significant difference between the reference feature amount and the feature amount of the image to be inspected, the surrounding particles are searched again for isolated particles (particles that are not associated with any), and Recalculate the amount and correct. When the perimeter or area shown in FIG. 3 is used as the feature amount, the feature amount can be recalculated by simply adding the feature amounts. FIG. 5 shows a flowchart of feature quantity comparison.

(Particle Search Range) FIG. 6 shows an explanation of the above-mentioned range for searching for isolated particles. In FIG. 6, (A) is set in a range in which each particle is surrounded by a circle,
(B) is set in a range surrounded by a rectangular area.

When the range of FIG. 6A is determined, it is determined so as to include the area occupied by each character. Therefore, even if a part of the character has a defect, it is compared with the recalculated feature amount within the search range of each particle, so that the amount of the defect can be accurately evaluated. In order to determine this range, when the reference image is input, contour tracing of each particle is performed, and settings are made so as to include each character as shown in FIGS. 6A and 6B.

Alternatively, as shown in FIG. 7, when the center-of-gravity distance to the particle around the pattern of interest is calculated and half the distance to the nearest particle is used as the basic radius, the entire particle that was originally one is It is advisable to set a distance that is considered to be sufficient, for example, a circular region having a radius of 2 to 3 times the basic radius.

[0016]

According to the mark inspection method of the present invention, the mark to be binarized has density unevenness and shading in illumination, and the unit of the independent pattern (particle) is the original unit in the binarization process. Even if it is different from the pattern, the defect of the mark can be accurately detected by comparing the feature amounts of the particles that correspond appropriately.

[Brief description of drawings]

FIG. 1 is a diagram showing a flowchart of the entire method of the present invention.

FIG. 2 is a diagram for explaining the feature amount of each particle.

FIG. 3 is a diagram showing a correspondence relationship of particles between a reference image and an inspection target image.

FIG. 4 is a diagram showing a correspondence relationship between particles when one particle is cut due to poor binarization.

FIG. 5 is a flowchart showing a feature quantity comparison method according to the present invention.

FIG. 6 is a diagram showing a search range of each particle.

FIG. 7 is a diagram showing a method of setting a range for searching for isolated particles when recalculating a feature amount.

Claims (1)

[Claims] 1. An input image signal is converted into a digital signal, and after binarization processing, a plurality of independent pattern feature amounts that are not connected to each other on a reference image are extracted to obtain a reference image and a reference image. In a mark inspection method for identifying a defect of a mark by comparing feature amounts of corresponding patterns with an inspection image, when the feature amounts of corresponding patterns differ by a preset allowable value or more, a reference image A mark inspection method characterized by comparing feature amounts in consideration of a pattern of an area occupied by each independent pattern therein.