激光雷达图像中的林区虫害区域分割与识别

Region segmentation and recognition of forest pests in image of laser radar

分割林区激光雷达图像区域,可实现虫害检测识别。传统方法使用图像同态模型分割算法对激光雷达图像林区虫害进行识别与分割,但精度不高,抗干扰性较差。基于虫害区域局域特征提取,提出激光雷达图像中林区虫害区域分割识别算法。首先利用高斯混合模型分析激光雷达图像的区域特征,构建局部区域平滑处理模型,提取局部林区虫害区域特征,实现激光雷达图像中林区虫害区域的分割与识别。仿真结果表明,该算法对激光雷达图像中的林区虫害区域分割识别精度较高,实时性较好,抗干扰能力较强,具有较好的应用性。

The pest can be detected by segmentated the laser radar image of forest. In view of the traditional homomorphic image of the Potts model segmentation algorithm of laser radar image forest pest identification sensitivity is not high, the problem is presented, which is based on laser radar image pixels on the edge adaptive contour feature extraction of forest pest region segmentation algorithm. Firstly Gaussian mixture model of laser radar image analysis of the characteristics of regional, and the local adaptive weighted control region segmentation and smoothness, edge pixel adaptive contour feature extraction, laser radar images of forest pest region adaptive segmentation was realized. The simulation results show that the algorithm can be used in the area of forest pests and insect pests in the forest region, which has high precision, good real-time performance, strong anti-interference ability, and has good applicability.