Study on the Method of Speed Detection and Image Recognition Based on Signal Normalized Cut Algorithm

In the image segmentation process, this paper uses the time series similarity principle to divide the image of track and field athletes into many similar blocks, and uses normalized processing to detect and identify the image, which greatly improves the speed and precision of image recognition. In order to verify the effectiveness and the reliability of the algorithm, this paper detects the arm speed of sprinter, finally the image segmentation normalized results and the curve of arm speed changing with time are obtained, which provides a new computer algorithm for identification and detection of arm speed signal.