The invention relates to a rail-mounted area extraction method based on laser point cloud data, and the method comprises the steps: carrying out the preprocessing of collected laser point cloud data in front of a train, and selecting the laser point cloud data in an effective range; the method comprises the following steps: by taking side walls on two sides of a rail running area parallel to a steel rail as reference objects, screening laser point cloud data based on a fixed distance segmentation method, clustering, and representing the laser point cloud data of the reference objects by a main cluster obtained after clustering; projecting the obtained laser point cloud data of the reference object to a horizontal plane, fitting a reference object curve based on an improved differential evolution algorithm, and taking a train left side reference object curve as an upper boundary and a train right side reference object curve as a lower boundary; according to the number of the laser point cloud data of the upper and lower boundaries obtained through fitting and a root mean square error, selecting a boundary line with a better index as a target boundary line based on a laser point cloud data volume-density two-step judgment method; calculating a rail-mounted area center line by taking the target boundary line as a reference; the height difference value of the upper boundary and the lower boundary serves as a judgment basis, a rail-mounted area boundary line expansion method or a rail-mounted area center line expansion method is selected and used in a segmented mode, and the rail-mounted area is calculated. Compared with the prior art, under the limitation of the collection performance of the laser radar, the method can extract a rail-mounted area with a longer distance, and facilitates the improvement of the obstacle detection distance and accuracy.
本发明涉及一种基于激光点云数据的轨行区域提取方法,包括:对采集的列车前方激光点云数据进行预处理,选取有效范围内的激光点云数据;以与钢轨平行的轨行区域两侧边墙作为参照物,用基于固定距离分段方法筛选激光点云数据并进行聚类,以聚类后得到的主簇代表参照物的激光点云数据;将所得参照物的激光点云数据投影至水平面,基于改进差分进化算法拟合参照物曲线,以列车左侧参照物曲线为上边界,列车右侧参照物曲线为下边界;依据拟合所得上、下边界的激光点云数据的数量、均方根误差,基于激光点云数据量‑密度两步判定法选取指标较优的一条作为目标边界线;以目标边界线为基准,计算轨行区域中心线;以上、下边界的高度差值作为判断依据,分段选择使用轨行区域边界线拓展方法或轨行区域中心线拓展方法,计算轨行区域。在激光雷达采集性能的限制下,与现有技术相比,本发明能提取更远距离的轨行区域,有助于提高障碍物检测距离和准确率。
Rail-mounted area extraction method based on laser point cloud data
一种基于激光点云数据的轨行区域提取方法
2023-11-03
Patent
Elektronische Ressource
Chinesisch
An improved iterative nearest point registration method for vehicle-mounted laser point cloud
British Library Conference Proceedings | 2023
|Rail-mounted area dispatching command system, device and medium
Europäisches Patentamt | 2023
|Europäisches Patentamt | 2023
|