基于单目视觉的静止目标定位方法

Location Method of Static Object Based on Monocular Vision

以无人机感知与避障为背景,提出了基于运动的单目视觉测距方法.选择摄像机运动过程中在不同位置对同一目标获取的两幅图像,然后利用尺度不变特征变换算法对所选图像进行特征检测和匹配,通过分析同一目标特征点在两幅图像中不同成像位置的变化,结合无人飞行器自身的运动参数,求解出无人飞行器与障碍目标之间的位置信息.采用该方法对不同位置的目标进行定位实验,结果表明该算法的测量精度和时效性可以满足实际避障的要求.

Aiming at the background of sense and avoiding of unmanned aerial vehicle,a monocular vision method for distance measurement from motion was proposed.Firstly,two different images were chose from a sequence of images for the same target in different position.Then,feature points could be extracted and matched through scale-invariant feature transform algorithm for the two images.Through analysis the different positions for the same feature point on different images and combined with the movement parameters of the aircraft,the information between the aircraft and the obstacle could be got.With the proposed method,the target was located at different locations.The experimental results show that the measurement accuracy and effectiveness of the method can meet the demand of avoidance.