Product quality inspection method

Abstract

The invention discloses a product quality detecting method. The method comprises two parts, wherein the first part is to construct a recognizer, and the second part is to use the recognizer to detectthe quality of a product to be tested, wherein the recognizer includes a dichotomic classifier which judges whether an object to be tested satisfies the requirements of products in a test category, aproduct detector which judges the position of the product in a picture, and a multi-classifier which judges whether the product is a qualified or a defective product. With the method adopted, whetherthe product is a qualified or defective product can be detected. The method has the advantages of high speed, extendibility and high flexibility.

Description

technical field [0001] The invention relates to a product quality inspection method, in particular to an expandable identification method for good and repeated products (defects, stains, whether or not a device) based on a small number of samples. Background technique [0002] In the industrial field, the need for identification of good and defective products is increasing. Defective products mainly include color difference, defects, deformation, etc. To give a simple example in life, if there are stains, incomplete, or relatively incomplete napkins in daily use, the stains may be It makes people even more disgusted, and there are defects in screws that can easily cause accidents. The above problems are all major problems that seriously affect the development of enterprises. [0003] The current existing methods are relatively single, mainly dealing with a single task, and a method can only do one task, such as only stain detection, or only defect detection, and the methods ...

AI Technical Summary

Problems solved by technology

[0002] In the industrial field, the need for identification of good and defective products is increasing. Defective products mainly include color difference, defects, deformation, etc. To give a simple example in life, if there are stains, incomplete, or relatively incomplete napkins in daily use, the stains may be It makes people even more disgusted, and there are defects in screws that can easily cause accidents. The above problems are major problems that seriously affect the development of enterprises.

[0003] The current existing methods are relatively single, mainly dealing with a single task, and a method can only do one task, such as only stain detection, or only defect detection, and the methods used are relatively traditional, scalable and flexible Poor, room for improvement

Images