用四叉树算法建立自主系统环境模型

Autonomou System Environment Modeling Based on Quadtrees

对环境建模是自主系统路径规划的基础,针对传统方法中存在的种种缺陷,该文在阐述自主系统中利用激光测距成像雷达的坐标变换法所获取的距离图实时检测障碍物的基础上,利用二值图像的四叉树表示是基于数据区域的一致性判别准则和空间递归分解原理建立起来的一种分层的树形数据结构,将一幅二值图像分为大小相同的四个分区,着重研究了利用模式四叉树建立部分已知或完全未知的环境模型的方法,结果显示其节省了四叉树占用的存储空间,在相同的存储空间情况下,提高了环境分辨率,加快了系统的运行速度,并给出了环境信息的优化策略。

Environment modeling is the basis for path planning of autonomous system, there is several faults in the conventional methods. A pattern quadtrees environment modeling study in this paper is proposed for partially known or completely unknown environment on the basis of describing real - time obstacle algorithm, coordinate transformation using range image of laser range sensor in autonomous system, the quadtrees of two value image composed of equal four area is the treely data structure divided up layer, which is built on the criterion of the data area in concert and the analytical principle to return progressively in space, the result show that it can save the memory of quadtrees,in the same memory, the environment identification can be improved and the system running is more quickly. In addition,a environment information optimizing strategy is given.