摘要:
激光成像雷达在工作时会受到噪声干扰,因此需要滤除距离像中无法反映真实信息的异常点。根据激光成像雷达距离像噪声的分布,推导出距离像中非异常值的分布特性,并提出了一种距离像异常值的判定条件。将判定条件表示为小范围内抽样分布的函数,利用稳健统计中的中位值法可估算出判定条件的取值,从而可以移除异常点得到噪声抑制后的距离像。利用该算法分别对激光雷达距离像进行仿真和实际获得的激光雷达距离像进行处理并进行数学分析,并与中值滤波进行了比较,结果表明基于统计特性的噪声抑制算法同时具有很强的噪声抑制能力和保护目标细节信息能力,可以满足激光成像雷达距离像噪声抑制的要求。算法充分利用了激光成像雷达较高的测距精度,可以应对多种场景下的需求,具有很强的实用性。
Abstract:
Ladar range image is getting more and more attentions in the field of target recognition. The ladar range image will be severely disrupted when the ladar is working in a poor condition, it is necessary to remove the abnormal points that can not reflect true information. The distribution characteristics of the normal points are derived based on the distribution of the noises in the ladar range image, and a judging criterion of the abnormal points in ladar range image is obtained. The criterion can be expressed as a function of sampling distribution in small area, and its value can be estimated by the median method of robust statistics. Hence the anomalies can be removed and the ladar range image after noise suppression can be obtained. Both the simulation analysis and the actual ladar range image processing are carried out based on the proposed arithmetic and compared with the median filter arithmetic. The results show the noise suppression arithmetic based on the statistical property of ladar range image can not only remove noises very effectively, but also protect the details well. The algorithm makes full use of the high ranging accuracy of laser imaging radar, has the ability to deal with the requirement of different scenarios, and it is proved to be strongly practicable.
点击查看大图
下载:
