Surface defect detection of aero-engine blades based on improved Faster-RCNN

Aiming at the problems of large surface noise and low detection accuracy in the process of aero-engine blade defect detection, a blade surface defect detection method based on improved Faster-RCNN is proposed. The aero-engine blade surface defect dataset is constructed through blade surface defect image acquisition and typical defect labeling. Through experimental comparison, VGG16 is selected as the feature extraction network; the CBAM attention mechanism module is combined in the backbone network, and the cosine annealing learning strategy is used to replace the original step-down learning strategy, so as to improve the extraction ability of the blade surface defect features. Comparative experimental results show that the proposed method improves the average accuracy by 3.45% compared with the Faster-RCNN algorithm.