Image super resolution by double dictionary learning and its application to tool wear monitoring in micro milling

摘要

It is an effective means to improve the machining quality of product by monitoring the tool wearconditions in micro milling. However, due to the small size and high speed of the micro millingcutter, it is difficult to capture the clear image of tool wear by CCD camera. Therefore, a doubledictionary sparse coding super-resolution (SR) method is proposed to improve the resolution ofmicro milling tool wear images in this work. Based on the traditional single dictionary, this workproposes the residual dictionary to reconstruct the residual image of the tool, achieving thedouble dictionary coefficient encoding reconstruction. In order to extract local features of lowresolution (LR) tool wear images, the Gobor filter banks are designed, and then the PCA method isused to reduce dimension of extracted features. The dictionary learning based on a single highresolution (HR) image is realized, which reduces the dependence of super-resolution algorithm onlarge datasets. The effectiveness of Gabor filter and double dictionary for restoration tool wearimage details is verified by experiments. Compared with proposed method, traditional interpolationmethod, sparse coding super resolution model 32 and super resolution convolutionalneural network model 26, the PSNR mean value of those methods are 43.62, 42.19, 43.20,43.21, respectively, which shows that the proposed method has excellent performance in superresolutionreconstruction of micro milling tool images for tool wear monitoring.