基于计算机视觉的飞机高度表示数识别

Manual reading is a key factor of constraining the automation and efficiency of altimeter detection system. The recognition based on computer vision has distinct advantage. This paper aims to the altimeter number of automatic recognition ,by combining the appearance of altimeter and handling of the dial grayscale images, filtering, thresholding, morphology pointer separation, thinning, and Hough transform to extract the pointer linear characteristics of image. The experiment shows that the algorithm has good performance. The slightly changing of the algorithm also appropriate to other pointer instrument recognition.