一种新的激光雷达距离像噪声抑制方法

主要针对激光雷达距离像的距离反常噪声抑制问题,阐述了激光雷达距离像的噪声原理,分析了应用传统中值滤波方法抑制距离反常噪声的缺陷,提出了基于包围准则的自适应中值滤波算法。该方法首先根据包围准则检测噪声,对5×5滤波窗口内的像素值进行排序差分;然后选择低于门限长度最长的连续差分值对应的像素值作为距离正常值;最后运用中值滤波和加权均值滤波进行噪声抑制。实验结果表明,该方法有效抑制了距离反常噪声,且较好地保护了距离图像中目标的边缘细节,均方根误差分别比3×3和5×5窗口中值滤波法减少了27.1%和9.1%。

A New Noise Suppression Method of Laser Radar Range Image

In order to suppress range anomalies noises of laser radar range image, the noise mechanism of laser radar range image is demonstrated in this paper. The shortcomings of traditional median filtering used to suppress range anomalies noises are analyzed. The range anomalies noise suppression method of range image based on surrounded guideline and self-adaptive median filtering is put forward. Firstly, the range anomalies noise is detected by surrounded guideline. The sorting and difference are carried out for pixels within 5 ~ 5 filtering window. Then, the pixels corresponding to continuous difference values of maximum length below threshold are regarded as range normal values. Finally, The weighted average method and self-adaptive median filtering are carried out for range anomalies pixel. Experimental results show that range anomalies noi- ses are suppressed efficiently by using this method. The edge and detail of object in range image are protected well. The root-mean-square-error（RMSE） is reduced by 27.1% and 9.1 ~ in comparison with median filtering methods of 3 × 3 and 5 × 5 filtering window. Key words laser radar; range image; range anomalies noises suppression; sorting-difference; image processing