基于RGB颜色模型的林区道路图像分割  

Image Segmentation of Forest Road Based on RGB Color Model

作　　者：王露露 Wang Lulu(College of Optical Electromechanical Engineering,Zhejiang A&F University,Zhejiang Hangzhou 311300)

机构地区：[1]浙江农林大学光机电工程学院,浙江杭州311300

出　　处：《南方农机》2021年第20期42-44,共3页

基　　金：浙江省基础公益研究计划项目(LGN18F030001)。

摘　　要：为了降低光照对林区道路的干扰,提高不同光照强度下道路的分割精度,将RGB道路图像的B分量分离出来获得灰度图,并通过Otsu对灰度图进行二值化。定量分析结果显示,精确率均值为0.977 5、召回率均值为0.996 9、F1均值为0.986 9,精度较高,鲁棒性较好,为实现林区道路分割提供了一种思路。In order to reduce the interference of light on forest roads and improve the segmentation accuracy of roads under different light intensities,the B component of the RGB road image is separated to obtain a grayscale image,and the grayscale image is binarized and quantitatively analyzed by Otsu.The quantitative analysis results show that the average precision rate is 0.977 5,the average recall rate is 0.996 9 and F1 is 0.986 9,it has high accuracy and good robustness and provides a way of thinking for realizing road segmentation in forest areas.

关 键 词：林区道路 图像分割 OTSU 颜色分量 

分 类 号：TP391.4[自动化与计算机技术—计算机应用技术]