Ultraviolet Image Processing Method in Corona Detection

摘要

Ultraviolet image detection technology is the first choice of the corona detection in power system, which could fulfill an all-weather reliable detection on corona. However, the solar-blind ultraviolet signal issued by corona discharge is relatively weak, and the quality of ultraviolet image is seriously affected by the noise from the detection system. The noise characteristics of the ultraviolet image that is detected by the corona has been analyzed in this paper, one kind of image denoising methods based on mathematics morphology and independence component analysis is put forward, and is applied to the corona and solar-blind ultraviolet image processing method. The method which improves the SNR of images, enhances the goal and weakens the noise is proved by experiments.