应用于紫外检测技术的ICA图像融合算法  

ICA image fusion algorithm applied in UV detection technology

作　　者：马立新[1] 张建宇[1] 周小波[1] 

机构地区：[1]上海理工大学光电信息与计算机工程学院,上海200093

出　　处：《机电工程》2016年第1期111-115,共5页Journal of Mechanical & Electrical Engineering

基　　金：上海张江国家自主创新重点资助项目(201310-PI-B2-008)

摘　　要：针对高压电气设备电晕放电故障点定位问题,对ICA图像融合算法、紫外检测技术和紫外成像系统等方面进行了研究。将ICA图像融合算法运用到高压设备电晕放电检测中,根据高压设备局部放电会发出紫外光的原理和紫外光波长在日盲200 nm^400 nm波段的特点,对高压设备局部放电区域采集紫外光图像和可见光图像进行图像融合,在融合图像中准确定位局放故障点。研究针对紫外图像和可见光图像灰度直方图差别大的特点,对传统Fast ICA算法进行优化,以优化后的权值对融合系数进行加权处理,建立融合规则使局放故障点表达更加清晰,并将算法运用于紫外成像系统。研究结果表明,改进后的Fast ICA算法对比于紫外成像仪的其他算法所得图像信息更加丰富、定位更加精准,融合图像可以有效地定位高压设备电晕放电故障点。Aiming at the phenomena of locate the point of high voltage corona discharge equipment fault, The ICA image fusion algorithm and ultraviolet detection technology, ultraviolet imaging system was studied to solve the problem. The ICA image fusion algorithm was used in high voltage corona discharge equipment testing, ultraviolet discharge image and visible image was acquired and integrated in the discharge area according to the ultraviolet wavelength is from 200 nm to 400 nm in sun-blind band. Then the accurate discharge point was located on fusion image. Research aiming at uv and visible light image＇s gray histogram were much different, traditional FastlCA algorithm and the weights of weighted fusion coefficient was optimized before processing, fusion rules was established to make partial discharge fault point ex- press more clear. Experimental results show that the algorithm compare to the rest of the uhraviolet imager＇s algorithms, the image informa- tion is richer, more accurate positioning and image fusion can effectively locate the high voltage corona discharge equipment point of failure.

关 键 词：紫外检测技术 图像融合 FASTICA算法 融合规则 

分 类 号：TM51[电气工程—电器] TP391.41[自动化与计算机技术—计算机应用技术]