Research on Infrared Image Processing Method of Dry Hollow Reactor

Infrared detection technology is currently an effective means for the power industry to grasp the operating state of the equipment in the state of non-power outage, which is widely used in the live detection profession of the power system, and has become an important means of detecting the operating state and abnormal fault of the dry hollow reactor. However, the shortcomings of infrared image blur and weak contrast are also important factors affecting the post-analysis of professionals, so it is necessary to process the infrared image of the dry hollow reactor taken at the scene. In this paper, the field-acquired infrared image is first grayscaled, and then the high-density impulse noise image based on the improved pixel peer group is used to remove the infrared image based on the characteristics of the dry hollow reactor itself, and the Retinex enhancement algorithm based on adaptive weight is used to enhance the infrared image. Through experiments, it is concluded that this method can reduce the influence of other factors on the image imaging effect, highlight the required parts and corresponding details, and facilitate the staff to accurately grasp the operating status and defects of the reactor in the next step.