IMAGE FUSION BASED ON MPCNN AND DWT IN PCB FAILURE DETECTION
Yiming Yuan, Ming Jiang, Wengen Gao
College of Electrical Engineering, Anhui Polytechnic University, Anhui, China
The traditional contact-type printed circuit board (PCB) test methods have been unable to meet the needs of the fault detection and maintenance of a variety of increasingly complex electronic equipment. The visible and infrared respectively reflects the background information and the radiation information of PCB, so we can fuse the visible image and infrared image of the board together, and use the new fusion image to locate and identify the abnormal high temperature components or areas of the circuit board. A novel fusion algorithm of multi-sensor image is proposed based on Discrete Wavelet transform (DWT) and pulse coupled neural networks (PCNN) in this paper. Firstly, the IR and visible images are decomposed by DWT, then a fusion rule in the DWT is given based on the PCNN. This algorithm uses the local entropy of wavelet coefficient in each frequency domain as the linking strength, then its value can be chosen adaptively. After processing PCNN with the adaptive linking strength, new fire mapping images are obtained. According to the fire mapping images, the firing time gradient maps are calculated and the fusion coefficients are decided by the compare-selection operator with firing time gradient maps. Finally, the fusion images are reconstructed by wavelet inverse transform. The proposed algorithm of image fusion using modified pulse coupled neural networks (MPCNN) and DWT results in better quality of fused image with Entropy, Average grads, Cross-Entropy as compared to conventional image fusion Algorithms.