摘要
为了研究逆投影变换方法在智能车驾驶环境信息采集中的应用,提出了一种反向求解逆投影的方法,即根据用户感兴趣的道路区域规划逆投影变换后的图像,并考虑了车道线的方向特性,提出分段的定向二维插值方法求解逆投影变换后图像各个像素点的灰度值。理论分析和标定试验表明,使用该方法进行逆投影变换后的图像能够较好地反映客观路面情况,并且可以降低逆投影变换的计算量,强化车道线信息。
A reverse inverse perspective transform （IPT） method was developed to analyze environmental information collected by intelligent vehicles. The image after the transform can be arranged according to the users＇ wishes. Directional 2-D interpolation in various subsections was used to calculate the pixel gray values to analyze the directional characteristics of the lane markers. Theoretical analyses and calibration tests show that with the directional 2-D interpolation based IPT, the image accurately represents the road environment with much fewer calculations and improved lane marker identification.
出处
《清华大学学报（自然科学版）》
EI
CAS
CSCD
北大核心
2006年第5期712-715,共4页
Journal of Tsinghua University(Science and Technology)
基金
国家"十五"科技攻关项目(2002BA404A21)
关键词
汽车
机器视觉
逆投影变换
定向二维插值
automobile
machine vision
inverse perspective transform
directional 2-D interpolation
