This paper studies the application of data fusion technology in power system to solve some difficult problems in this complex energy system. A transmission tower identification and bird nest detection method based on corner, line, color and shape features is proposed. Through LSD (Line Segment Detection) and Harris corner detection method, the straight line segment and corner point in the image are extracted respectively. Combined with triangle method, the actual tilt angle of tower is measured; According to the nesting rule of birds in transmission towers, the basic unit segmentation algorithm of transmission towers is proposed, and the basic unit segmentation of transmission towers is realized by using the local maximum of the target pixel row statistical histogram. The algorithm proposed in this paper can effectively solve the problems of on-line measurement of tilt angle of transmission tower and on-line detection of bird's nest, which will lay a theoretical foundation for on-line monitoring of transmission tower status.