【摘要】 面向自动驾驶技术,研究一种实时的道路边缘检测方法—MERD(Mobile Equipment based Road-edge Detection).该方法基于智能移动终端的Android系统平台,将针对HSV图像的数学形态学方法和针对灰度图的Hough变换检测方法相结合,针对不同路段环境进行道路边缘检测.实验验证了该方法在处理傍晚路段、阴影较多路段等不同情况的道路环境时,能够提供较高准确度的实时道路边缘检测,为车辆的自动驾驶提供安全保障.更多还原

【Abstract】 This paper proposed a real-time method for road-edge detection. The method is named as the mobile equipment based road-edge detection(MERD). Based on the Android platform in the intelligent mobile equipment, MERD combined the mathematical morphology based detection in HSV image with the Hough transform based detection in gray scale image.MERD can be used to provide road-edge detection for various road situations. Experiments are given to show the validity and accuracy of MERD in real-time road-edge detection for the shadowy road, road in the evening, road with snow, road with rain, road with pedestrian and shadows, thus providing the safety insurance for self-driving vehicles.更多还原