Research on infrared image segmentation and fusion of substation based on modified unit‐linking‐pulse coupled neural networks and affine speeded up robust feature

Abstract

Abstract Aiming at the problem of fault point temperature reading and positioning, target device segmentation, and fault point matching fusion in infrared warning system of substation inspection robot, MUL‐PCNN and ASURF are applied to realizing accurate fusion between the fault area in infrared image and visual image. The experimental results show that this method has certain value in the aspect of infrared image segmentation and fusion of substation.