摘要
针对在露天矿区无人驾驶卡车道路识别中,受矿区道路边缘模糊、阴影与车辙印干扰等因素的影响,往往无法准确检测到实际道路的问题,提出一种基于HSV空间图像处理的矿区道路边缘检测方法。在HSV颜色空间内,利用半阈值小波对图像进行初步去噪,针对矿区道路边界模糊、边缘退化等问题,提出一种基于融合策略的道路图像增强分割处理算法,可以有效地突出道路边缘,最后利用Canny算子实现对矿区非结构化道路边缘检测。实验结果表明,该方法的检测准确率可达到90.0%以上,能有效识别复杂背景下矿区非结构化道路边缘。
To solve the problem that the road edge of unmanned truck can’t be effectively identified when it was travelling,due to the reasons of the unstructured roads designing and the complicated background.In this paper,we proposed a method of road edge detection for mining,based on the HSV spatial image recognition.The edge features of the mining road are effectively highlighted by the noise reduction processing and image enhance mental processors with each component of the HSV space.By the way,combining with road segmentation,the edge of unstructured road in mining area will be detected,using Canny operator.The experimental result shows that the accuracy of this method for detecting unstructured roads in mining areas can be above 90%,and the edges of unstructured roads in mining areas can effectively identified under complex background.
作者
卢才武
马玲
阮顺领
LU Caiwu;MA Ling;RUAN Shunling(School of Management,Xi'an University of Architectural Science and Technology,Xi'an 710055,China)
出处
《测绘科学》
CSCD
北大核心
2020年第9期123-131,共9页
Science of Surveying and Mapping
基金
陕西省自然科学基础研究计划项目(2019JM-492)。
