Optimization method of Cognex surface quality detection system of hot galvanizing production line

The invention belongs to the technical field of steelmaking, and discloses an optimization method of a Cognex surface quality detection system of a hot galvanizing production line. The optimization method comprises the following steps: adopting a Smartview to construct a system server database in an in-situ industrial personal computer; adopting two CCD (Charge Coupled Device) line scanning cameras to simultaneously detect a surface; through a specific defect filtering operation, filtering defects with small sizes; through a direct judgment operation, directly judging the defects with a big size; adopting a way that a picture shot by real-time surface detection is compared with a real object morphology to establish a defect map depot, wherein multiple pieces of defect pictures are adopted for each defect in the defect map depot, and a self-learning feedback control algorithm is adopted to perfect the defect depot; and removing the pictures of which the self-learning accuracy is lower than 90%, and supplementing new pictures to carry out self-learning again until the accuracy is higher than 90%. The defection automation degree and the detection efficiency and reliability of band steel surface defects can be greatly improved.