环境污染区域测绘数据点云分类精简方法研究

Research on Point Cloud Classification and Simplification Method for Surveying and Mapping Data of Environmental Pollution Areas

针对环境污染区域的测绘点云数据由于具备随机性、模糊性特点,导致点云数据分类不准确的问题,提出一种基于点云分类器的环境污染区域测绘数据点云分类精简方法。在KNN点云采样算法中,设置k值划分数据,使用曼哈顿距离作为划分的判别标准,均衡采样密度,通过KNN算子提取独立点特征,利用支持向量机作为点云数据的分类器,实现环境污染区域测绘点云数据的精简分类。测试结果表明,在5种不同程度的污染等级下,设计的分类方法的各个指标均优于两种传统的分类方法,验证了设计方法在实际应用中的有效性。

Aiming at the problem of inaccurate classification of point cloud data due to the randomness and fuzziness of mapping point cloud data in environmental pollution areas,a simplified method of point cloud classification based on point cloud classifier was proposed.In the KNN point cloud sampling algorithm,k value is set to divide the data,Manhattan distance is used as the discrimination standard for division,sampling density is balanced,independent point features are extracted by KNN operator,and support vector machine is used as the classifier of point cloud data to achieve simplified classification of the mapping point cloud data in environmental pollution areas.The test results show that under 5 different pollution levels,each index of the designed classification method is superior to the two traditional classification methods,which verifies the effectiveness of the design method in practical application.