RETRACTED ARTICLE: Capacitance pin defect detection based on deep learning

Mask R-CNN network based on deep learning algorithm is specifically optimized for the small visual defects of gears. After comparison, the ResNet-101 residual neural network is used as the image sharing feature for network extraction. Subsequently, the unreasonable convolution of the feature extraction process P_5 in the characteristic pyramid network is removed to improve the defect detection rate indicator. Finally, the sizes of the anchor and label frames are adjusted according to the dimensioning of the tiny objects in the capacitance sample, and an appropriate aspect ratio is set to achieve the effective training of the network in the candidate area. Experiments show that the optimized Mask R-CNN network can achieve a defect detection rate that is above 98