According to the idea of depth of field map blurs background

a video speed measurement algorithm based on depth of field and LK optical flow method is proposed.The multi-scale adaptive wavelet algorithm is used to decompose the image in four directions:horizontal

vertical and diagonal; the edge map and the Li’s index are introduced to optimize the depth of field map generated by the relationship between the high-frequency wavelet coefficients and the depth values of the salient feature regions; The LK optical flow method is used to detect the ORB corner point of the depth of field map

and the SUSAN algorithm is used to eliminate the abnormal corner points and draw the corner point distribution map.The movement speed and angle of the target are calculated from the displacement of the corner point coordinates in the adjacent frame image.The experimental results show that the algorithm can effectively suppress background influence

reduce the amount of calculation

and improve the precision and real-time performance of moving targets in complex background and dynamic background.