Research on the Identification Method of Colour-Based Status Indicator Based on Feature Matching

An indicator light recognition algorithm based on feature matching is proposed to solve the problem of robust recognition of colour status indicators in indoor substation cabinets. To eliminate the effect of image distortion, an indicator region recognition method based on feature point matching is proposed, which uses a perspective transformation matrix to realize the state transformation from the detected image to the template image. The algorithm can detect the brightness anomaly of the image and eliminate the effect of brightness anomaly by gamma correction; the indicator status is finally judged by two matching criteria. Image recognition experiments are conducted using image information of indoor substation cabinets, and the experimental results show that the colour-based status indicator recognition algorithm based on feature matching is able to accurately and robustly determine the indicator status.