Automatic production and automatic learning method of machine vision system detection algorithm

Abstract

The invention relates to an automatic production and automatic learning method of a machine vision system detection algorithm. The method is characterized by comprising the following steps of: (1) establishment of a target structure model knowledge base, analyzing a detection point and a detection method aiming at the type of a specific detection target, and then selecting a corresponding detection algorithm according to detection areas decomposed by the target structure model; (2) an automatic learning process of discriminant parameters of the detection algorithm: after the step (1) is finished, automatically establishing high and low threshold values of the discriminant parameters of the detection algorithm so as to realize the automatic learning process; (3) the automatic correction ofalgorithm parameters: after a system begins to carry out on-line detection, along with the operation of the system, reserving a historical image of a section of an event by the system, and calculating the detection parameter of the reserved historical image; correcting the original high and low threshold values according to the currently calculated detection parameter while using the corrected high and low threshold values as a new detection standard so as to realize the automatic production and automatic learning method of the machine vision system detection algorithm.

Description

technical field [0001] The invention relates to the technical fields of computer vision and image processing, in particular to an automatic generation and automatic learning method of a machine vision system detection algorithm when using computer image processing technology to realize online detection of target appearance quality. Background technique [0002] With the development of modern industry, the online automatic detection technology belonging to the automation industry has been playing a powerful role in reducing the intensity of manual labor, replacing manual detection to liberate labor, and improving labor productivity. It has achieved unprecedented development and is widely used in electronics, Working condition monitoring, finished product inspection and quality control in tobacco, pharmaceutical, printing, food, automobile and other industries. [0003] According to statistics, in 2006, the sales of the German machine vision industry increased by 9 percentage ...

AI Technical Summary

Problems solved by technology

[0004] Machine vision Since the 1980s, the machine vision that has been implemented has been in the state of "one third can be used, one third has failed, and one third can be used after modification". Most of the enterprises fail, and even the media claims that "the success rate of machine vision implementation is zero". The cost exceeds the estimated cost by 178%, the installation time exceeds the estimated time by 230%, and the company's loss rate after use has reached an astonishing 59%. However, there are still more than 20,000 companies in the world investing more than 10 billion US dollars to adopt Machine vision system, and there is an upsurge of machine vision in China's large and medium-sized enterprises. They also use machine vision, and the difference between the results is like heaven and hell. The most important reason is the stability and stability of machine vision equipment. Ease of use. Originally, computer vision and image processing technology are relatively professional. It is difficult for skilled workers in the production line to get familiar with machine vision quickly. If the equipment needs a lot of maintenance work after it is put into use, the equipment will be difficult to be accepted. and successful application, but the current machine vision (including the so-called smart camera) all adopt the architecture: professional computer image processing algorithms are provided to the user, and the user sets the corresponding algorithm on the imaged image according to the detection content he thinks , if the detection is unreliable during use, the user will modify the detection algorithm according to the new image. If there is a need for a new detection grade, then set a new grade and configure a new algorithm. This architecture makes the detection technology unable to achieve Stability and ease of use, cannot be used as easily as traditional sensors