|
计算机应用研究 2011
Vehicle-based single-line laser radar image segmentation and noise detection algorithm
|
Abstract:
Laser radar has more valuable advantages with fine resolution capability and the possibility of day\night operation in all weather, so it becomes a hot research on intelligent vehicle. The laser radar is usually installed on the front bumper in order to detect the obstacles in front. After analyzed laser radar image segmentation algorithm at home and abroad, an improved algorithm based on ABD (Adaptive Breakpoint Detector) is proposed, the experiments show that the new segmentation algorithm is more rational and effective. And an efficient method for noise point detection based on the improved segmentation algorithm is also proposed, the results are ideal. The single-line laser radar would detect the road surface, but actually it is not a real obstacle, a local and entire region combining algorithm based on the tremendous difference of current frame and previous frame is proposed. Experiment results show that the road detection is effective. The research mentioned above is valuable for vehicle sensor data fusion.