Research On Dual CCD Image Monitoring Method And Its Application In Substation Inspection AGV

With the development of Smart Grid Substation and the increase of voltage level, substation area is gradually expanding, which brings a great challenge to the electrical equipment inspection work station. The traditional inspection methods rely heavily on the experience of technical staff, sensory and handheld test equipment, which can not meet the growing automation and intelligent demands of the substation inspection work under the new situation. If we put power robot of image processing function into the substation inspection and monitoring, it not only can work in all weather, but also can greatly assist the staff of electrical equipment for reliable inspection, effectively avoiding of the risk and reducinglosses.The image monitoring technology equipped with double CCD inspection robot were related research, image matching with visible light and infrared image fusion technique are studied, and based on DSP+FPGA architecture we design and develop the image processing system. The system carrying conventional visible light CCD and thermal infrared camera, it can finish real-time matching and recognition for electrical equipment images and the equipment type template library collection; for those prone to overheating fault of electrical equipment, it can make a temperature monitoring with the implementation of infrared image fusion technology.This dissertation mainly introduces the key technology of image monitoring system, including CCD visible light image matching and infrared fusion and the hardware system and software system architecture based on CCD and DSP+FPGA. For the image matching technology, the specific methods of region matching and feature matching and their advantages and disadvantages are studied. On the basis of this, an improved matching method is proposed for the combination of the region and the feature information. Through the simulation experiment on Matlab platform, the improved algorithm is proved to have good performance and strong robustness. For the image fusion technology, this dissertation studies the image fusion based on wavelet transform and Nonsubsampled Contourlet transform (NSCT) framework. According to the calculation of NSCT fusion algorithm has high complexity; this dissertation proposes an improved algorithm. This dissertation firstly introduced the signal of the compressed sensing theory and applies it to high frequency coefficients of NSCT image, then used wavelet transform fusion algorithm for processing the low frequency coefficients and finally recoverd the image data the reconstruction algorithms. Through the Matlab simulation experiments,it show that the improved image fusion algorithm not only has good fusion effect, but also take into account the real-time needs of image monitoring system. The design of double CCD image acquisition, image matching, fusion and display module hardware circuit, completed the corresponding module of the DSP program, written in transplantation and optimization. The substation electrical equipment images collected in real experimental results show that this method can get satisfactory results.