System and method for detecting painted housing of white household appliance based on deep learning

Abstract

In order to upgrade the detection of white household appliances, the invention provides an intelligent detection unit based on deep learning, which is mainly characterized by designing a set of intelligent detection line for white household appliances and a detection method based on tandem semantic segmentation and a classification neural network. The intelligent detection unit for a white household appliance comprises a transport module, a stay module, an information collection module, and a processing analysis module. The transport module and the stay module cooperate electrically such thata detected object reaches a specified location, the information collection module is an industrial camera for acquiring an image of a detected area of the detected object, and the processing analysismodule obtains the type, location and size of a defect through tandem semantic segmentation and neural network classification. Through the cooperation of the modules, the physical quality informationis converted into detection information output by a neural network algorithm, which can effectively solve the intelligent detection problem of the painting quality of the white household appliance.

Description

technical field [0001] The invention relates to the field of intelligent quality detection of product appearance, in particular to an intelligent quality detection system and method for spray paint of white home appliance shells. Background technique [0002] White home appliances and other metal shells usually use painting technology to carry out shell packaging, anti-corrosion and product color personalization. People have high requirements for the appearance of the shell, but the painting technology is affected by machine precision and paint performance. defect. There are mainly defect types such as black spots, spray leakage, sagging, orange peel and wavy lines. When these defects are serious, the metal shell will be corroded, and when they are light, they will affect the appearance of the product. If they are not detected and fed back to the painting process in time, the enterprise Larger production accidents occur. [0003] Therefore, it is necessary to carry out qua...

Images