JP2006308535A - Selection device for learning region - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To enable failure substrates to be used for learning, and to significantly improve the operability of a defect detector. <P>SOLUTION: An image of the substrate imaged by a CCD camera 6 is sent to a personal computer 5, and a part of a defect is analyzed by a CPU 2 in the personal computer 5 and is displayed on a monitor 1. Although the displayed image of the defect accidentally includes a part, having no defect because the learning is not completed, operator needs only to enclose the part of the actual defect with a mouse 4, and information on the defective part is provided to the CPU 2. Based on the information, the CPU 4 removes the defective part and learns only the non-defective part. Thus, learning with the defective substrate can be made. <P>COPYRIGHT: (C)2007,JPO&INPIT

Description

Detailed Description of the Invention

[Technical field to which the invention belongs]
The present invention relates to a defect detection apparatus that detects defects on a printed circuit board, a lead frame, a printed material, and the like by a non-defective product learning method.

[Conventional technology]
In a defect detection device using a conventional non-defective product learning method, such as a printed circuit board defect detection device, it is necessary to prepare several non-defective substrates, but when the non-defective substrate cannot be prepared, the device cannot be operated. It was.

[Problems to be solved by the invention]
In a defect detection apparatus such as a printed circuit board that employs a non-defective product learning method, it is actually difficult and problematic to prepare a large number of non-defective substrates on the entire surface of the substrate. The present invention has been made to solve this problem.

[Means for solving problems]
A means for solving the above problem is that learning can be performed even on a defective substrate.
Specifically, it is designed to be able to learn by removing the defective part from the substrate. By doing so, even a defective substrate can be used for learning, and the operability of the defect detection device is remarkably improved. Improved.

[Mode for Invention]
FIG. 1 shows an example of the present invention. A monitor (1) for displaying an image of a substrate, a personal computer (5) incorporating a CPU (2), a memory (3), etc., and an image of the substrate are taken. A CCD camera (6) and a mouse (4) for drawing an area on the monitor.
The image of the substrate imaged by the CCD camera (6) is sent to the personal computer (5), where the defect location is analyzed by the CPU (2) in the personal computer (5) and displayed on the monitor (1). . Since the displayed image of the defect has not been learned yet, a part that is not a defect is also displayed by mistake. However, the operator surrounds only the actual defect part with the mouse (4), and the defect part information is stored in the CPU. (2) Give. Based on this information, the CPU (4) excludes defective parts and learns only good parts. In this way, learning with a defective substrate becomes possible.

[The invention's effect]
By using the present invention, a defective substrate can be used for learning, and the operability of the defect detection apparatus is remarkably improved.

1 illustrates an example of a configuration of the present invention.

Explanation of symbols

1 is monitor 2 is CPU
3 is memory 4 is mouse 5 is personal computer 6 is CCD camera

Claims (1)

A defect detection apparatus having a function of learning by removing a defective portion in a defect detection apparatus that detects an image by capturing an image with an imaging element such as a CCD camera and analyzing the captured image by a non-defective learning method.

Images