一种基于局部聚类的激光雷达距离像噪声抑制方法
首发时间:2021-01-08
摘要:激光雷达图像具有广泛的应用场景,广泛应用于测距、探测、地理绘制、激光制导等领域。为了解决在获取的激光雷达距离像存在噪声的问题,提出了一种基于局部聚类的噪声抑制算法,该方法将邻域的距离值排序后进行聚类,若当前点位于判定为噪声的聚类中则将其置为最大聚类的中值。然后,将此方法和中值滤波以及中值加均值的滤波算法在仿真集和实际激光雷达图像上进行了对比。结果显示,该方法能够有效抑制距离反常噪声,相比较中值滤波和中值加均值的滤波算法,其引起的失真较小,处理后的噪声点数少,并且能保留一定的细节。
For information in English, please click here
A LiDAR range image noise suppression method based on local clustering
Abstract:Lidar image has a wide range of application scenarios, widely used in ranging, detection, geographic mapping, laser guidance and other fields.In order to solve the problem of noise in the acquired LIDAR range image, a noise suppression algorithm based on local clustering was proposed, which sorted the distance values of the neighborhood and set the value of the current point in the determined noise cluster as the median of the maximum cluster.At the same time, this method is compared with median filtering and median plus mean filtering algorithm in the simulation set and the actual LIDAR image.The results show that this method can effectively suppress the abnormal distance noise. Compared with the median filtering and median plus mean filtering algorithm, the distortion caused by this method is less, the number of noise after processing is less, and some details can be retained.
Keywords: LiDAR image processing distance image noise suppression local clustering
基金:
引用
No.****
动态公开评议
共计0人参与
勘误表
一种基于局部聚类的激光雷达距离像噪声抑制方法
评论
全部评论0/1000