Abstract: In order to solve the problem of heavy workload and human error in industrial field pointer instrument identification，a method of industrial linear pointer instrument identification combined perspective transformation with Hough transform is proposed． After the image denoising and perspective transformation，Hough transform is used to detect the image features of the instrument pointer，and the actual pointer deflection angle is simulated with this image feature． Finally，the final number is calculated by this angle value，linear pointer instrument identification work is completed． When the frontal image cannot be well acquired in the field，recognizing industrial linear pointer meter readings is achieved well on some ways． In the new situation，a new way of thinking for the application of artificial intelligence in the industrial field is provided by this method．

Key words: image filtering, Hough transform, perspective transformation, computer vision