基于梯度图像法的激光雷达图像融合滤波算法  

A Fusing Filtering Algorithm of Lidar Image Based on Grads Image Method

作　　者：夏朝辉[1] 孟飞[1] 王仕成[1] 张合新[1] 

机构地区：[1]第二炮兵工程学院自动控制系,西安710025

出　　处：《兵工自动化》2012年第8期55-56,61,共3页Ordnance Industry Automation

摘　　要：为了有效消除激光雷达图像斑点噪声的干扰,针对单一滤波算法的不足,提出一种基于梯度图像法的激光雷达图像滤波融合算法。分析激光雷达图像的噪声特点,综合Lee滤波和小波滤波的优势,该算法利用图像的梯度图像来判断斑点噪声隶属于信号区域还是边缘区域,在边缘点采用小波阈值滤波的结果,在信号区域选择Lee算法的结果达到最好的滤波目的。仿真试验的结果表明:该算法具有很强的散斑噪声抑制能力,能够在噪声滤波和图像信息保持2个方面都能达到很好的效果。In order to effectively eliminate the speckle noise interference of the laser radar image, it is difficult to gain better result by a single filter algorithm, and a fusing filtering algorithm of laser radar image based on grads image method was proposed. After the character of the noise of the radar image was analyzed and the advantages of Lee filtering and wavelet filtering was integrated, the image gradient image was used to determine the point belongs to edge region or to the signal region. The result of wavelet filtering was adopted if the point in the edge region, otherwise the result of Lee filtering was selected to achieve the best filtering effect. The simulation results show that the algorithm has strong speckle noise suppression capability, and it can obtain better effect on both noise filtering and image information holding when the new algorithm is used.

关 键 词：梯度图像法 激光雷达图像 融合滤波 Lee滤波 小波滤波 

分 类 号：TJ02[兵器科学与技术—兵器发射理论与技术]