Steel product surface defect detection method based on model compression

Abstract

The invention discloses a steel product surface defect detection method based on model compression, and belongs to the technical field of steel surface defect detection, and the method comprises the following steps: S1, preprocessing a hot rolled strip steel surface defect image sample; s2, building a hot rolled strip steel surface defect detection network based on an anchor-free target detection framework; s3, carrying out model compression on the defect detection network, and training a model by adopting a linear strategy and a training sample; and S4, testing the defect detection network on the independent verification set and outputting a detection speed and a detection result. According to the invention, the steel surface defect detection network for identifying and positioning is firstly established, and then redundant parameters in the defect detection network are removed, so that the high storage space of the model and the consumption of computing resources are reduced, and the defect detection network is more stably and efficiently landed on a hardware platform.

Description

technical field [0001] The invention relates to the technical field of steel surface defect detection, in particular to a method for detecting surface defects of steel products based on model compression. Background technique [0002] Cosmetic imperfections in a product not only spoil aesthetics and comfort, but can also seriously impair performance. Therefore, the surface defect detection of products must cover the intermediate joints of production and the last link before leaving the factory. Taking the detection of strip steel as an example, strip steel is the main material for the iron core of motors, generators and transformers. Surface defects such as cracks, inclusions, plaques, rolling scales, and scratches will seriously damage the corrosion resistance of strip steel. Anti-fatigue and iron loss characteristics directly affect the performance of motors, generators and other products. If the defective strip steel cannot be identified before leaving the factory, it m...

AI Technical Summary

Problems solved by technology

[0005] The technical problem to be solved by the present invention is: how to reduce the large amount of storage space and computing resource consumption of the model, and deploy the defect detection network on the hardware platform more stably and efficiently, so as to greatly reduce the detection cost while taking into account the detection accuracy, providing a method for surface defect detection of steel products based on model compression

Images