An Improved U-Net Network and Its Application in Gear Pitting Measurement

Abstract

Aiming at the problems of low segmentation accuracy of small targets, high computational complexity in U-Net, a new U-Net network based on dilated convolution and reconstructed sampling units (DSU-Net) is proposed. In DSU-Net, in order to increase the receptive field of image feature extraction and fuse multi-scale information, dilated convolutional layers with different dilation rates are designed; in view of the shortcoming of losing a large amount of semantic information during the pooling process, sampling units that combine pooling and convolution are constructed, and depthwise separable convolution is used for feature extraction, thereby enhancing the feature extraction capability of neural network and reducing the computational cost. The experimental results of Gear Pitting dataset show that DSU-Net has better segmentation performance than U-Net, ResU-Net and R2U-Net on the three metrics of IoU, Dice Coeff and F1 Score. The proposed method can calculate the gear pitting area ratio more accurately, so as to solve the problem of efficiently and accurately detecting gear failure in the gear contact fatigue test.