In order to solve the problem of low accuracy and efficiency in decoding the circular coded targets in close range photogrammetry, an efficient extraction and robust recognition algorithm for circular coded targets is proposed. In this algorithm, the noise and interference are filtered by the contour perimeter and the ratio of long axis and short axis after the least square ellipse fitting of the contour. In order to locate the coded targets accurately, the gray squared weighted centroid method

more »

... is taken to locate the centroid of coded targets. Finally, a new algorithm based on the decoding endpoint is proposed to decode the coded targets. Experimental results show that the average recognition accuracy can reach 99.2% when the angle between the camera optical axis and the normal of coded target is less than 60 • . Furthermore, the recognition accuracy rate can still reach 93% when the angle is 70 • . When applied to the images containing complicated background information, the global recognition accuracy, false recognition rates and the average elapsed time are 97.2%, 1.11% and 2.47s respectively, which validates the optimality of its comprehensive performance compared with other existing recognition algorithms.