陈云赛, 杨磊, 宫垠, 庞永杰, 沈鹏, 刘坤. 基于线结构光的水下目标扫描定位方法[J]. 机器人, 2018, 40(6): 786-793,802. DOI: 10.13973/j.cnki.robot.180158
引用本文: 陈云赛, 杨磊, 宫垠, 庞永杰, 沈鹏, 刘坤. 基于线结构光的水下目标扫描定位方法[J]. 机器人, 2018, 40(6): 786-793,802. DOI: 10.13973/j.cnki.robot.180158
CHEN Yunsai, YANG Lei, GONG Yin, PANG Yongjie, SHEN Peng, LIU Kun. Underwater Target Scanning and Location Method Based on Line Structured Light[J]. ROBOT, 2018, 40(6): 786-793,802. DOI: 10.13973/j.cnki.robot.180158
Citation: CHEN Yunsai, YANG Lei, GONG Yin, PANG Yongjie, SHEN Peng, LIU Kun. Underwater Target Scanning and Location Method Based on Line Structured Light[J]. ROBOT, 2018, 40(6): 786-793,802. DOI: 10.13973/j.cnki.robot.180158

基于线结构光的水下目标扫描定位方法

Underwater Target Scanning and Location Method Based on Line Structured Light

  • 摘要: 当海底环境噪声及声学散射影响比较严重时,声呐方法无法实现高精度的海底探测.因此,设计了一种基于线结构光的扫测系统.激光中心线条纹提取及系统参数快速标定是影响线结构光测扫精度的2大关键问题.在激光中心条纹处理方面,多阈值算法通过采用不同的阈值可以有效地滤除背景噪声,基于极值法的高斯拟合法则对部分激光条纹采用高斯函数进行拟合,结合多阈值算法和基于极值法的高斯拟合法提出了一个条纹中心线提取算法.该算法在保留多阈值法快速性的同时,又具备高斯拟合法的准确性.而在系统参数标定方面,提出了一种共线点的快速标定法,设计的平面标定靶只需通过确定3个共线的特征点即可快速实现系统参数的标定和高精度的海底快速测扫.最后,在水池中进行了模型实验,验证了该方法的可行性和有效性.

     

    Abstract: The acoustic method can't realize high-accuracy seabed survey when the seabed environment noise and acoustic scattering influence are serious, so a scanning system based on line structured light is proposed. Generally, the laser center stripe extraction and the system parameter rapid calibration are the two key problems affecting the accuracy of line structured light sweeping system. In the laser center stripe processing, the multi-threshold algorithm can effectively filter out background noise by using different thresholds, and the Gauss fitting method based on the extreme value method is used to fit part of the laser stripe with Gauss function. A stripe centerline extraction algorithm is proposed by combining the multi-threshold algorithm with the Gauss fitting method based on the extreme value method. The algorithm has the accuracy of the Gauss fitting method while retaining the speed of the multi-threshold algorithm. For the parameter calibration of the system, a fast calibration method is proposed for collinear points. With the designed plane calibration target, the system parameters can be calibrated quickly and the seafloor can be scanned with high precision by determining only 3 collinear feature points. Finally, a model test is carried out in water tank to verify the feasibility and effectiveness of the proposed method.

     

/

返回文章
返回