WEI Wu, GONG Shu-feng. Research on unstructured road detection algorithm based on machine vision[J]. 2010, 29(1): 39-41. DOI: 10.13873/j.1000-97872010.01.023.
基于机器视觉的非结构化道路检测算法研究
摘要
提出了一种提高非结构化道路边缘检测实时性和准确性的算法。该算法中首先对原始道路图像进行中值滤波
抑制随机噪声
并选择基于双峰法的多阈值Otsu方法进行图像分割
使分割效果和分割时间得到优化。通过Canny算子进行初次边缘检测
然后
采用数学形态学修正
从而得到完整、清晰的道路边缘图像。仿真结果表明:该非结构化道路检测方法具有良好的实时性和准确性。
Abstract
A kind of unstructured road detection algorithm which has improved real-time detection and accuracy is introduced.First
the original images are processed by median-filtering
and random noise is suppressed
and an Otsu multi-threshold algorithm is chosen based on two-peak method for rapid image segmentation
the segmentation effect and the segmentation time are optimized;Finally the primary edge detection is carried on with the Canny operator
and mathematics morphology is used to revise
thus obtain the complete and clear path edge image.The simulation result indicates that this unstructured path examination method has good characteristics of timeless and accuracy.