Signal detection of laser radar based on the background character parameter

The detection of the faint target in the lidar is one of the key technologies of the lidar. One of its difficulties is that the characteristic distinguishing the target and the noise is very lacking. The target this paper researches is the plane. After researching the practical echo signal, we find that the target often is the isolated dot in the echo signal. The target is irrelevant with the noise. However, the noise is expressed with the neighboring dot. To resolve the problem of the target detection in low signal-to-noise, an algorithm of the target detection based on the background characteristic parameter is presented. The clutter is processed using the background characteristic parameter of mean variance. In the small region, the mean variance of the background is steady. However, the mean variance of the target is distinct. Therefore, the signal-to-noise is improved greatly. Then with CFAR detection and multi-frame relative detection, the true target is captured. The experimentation had proved this algorithm improved the performance of the lidar. This algorithm is effective and does well on real time, it is valuable in practice.