摘要
为了优化苹果采摘机器人采摘路径,在获得苹果树场景三维位置信息的基础上,提出一种具有多种地形损耗的A*算法。结合ToF(Time-of-Flight)深度相机和Hu不变矩获取苹果和不同障碍物的三维位置信息,建立存在果实和多种障碍物的二维地图。在二维地图上,利用具有不同地形损耗函数的A*算法进行仿真实验。改进后的A*算法将障碍物分为可通过的障碍物(树叶)和不可通过的障碍物(树枝),且障碍物存在位置处的自带移动耗费向周围以线性递减,避免了基本的A*算法中只具有单种障碍物问题,从而增加了不同种类障碍物对路径选择的影响程度,优化了路径质量。对比实验表明:改进后的算法提高了对于复杂地图的处理能力,产生的路径长度更短,转折次数更少。
In order to optimize the picking path of the apple picking robot,this paper presents a A* algorithm with a variety of terrain loss on the basis of obtaining the 3 D location information of apple trees. First,ToF(Time-of-Flight)depth camera and Hu invariant moment are used to obtain the three-dimensional position information of apple and different obstacles,and a two-dimensional map of the existence of fruit and different obstacles is established. The A* algorithm with different terrain loss functions is used to simulate on the two-dimensional map. The improved A* algorithm divides the obstacles into the barrier(leaves)and the barrier(branch)that can not be passed,and the cost of self movement at the location of the obstacle decreases linearly,avoiding the problem of only a single type of obstacle in the basic A* algorithm. Thus,the influence degree of obstacles on route selection is increased,and the path quality is optimized. The experimental results show that the improved algorithm improves the processing ability for complex maps,and the new algorithm produces shorter path length and fewer turning.
引文
[1]宋哲,王宏,里程辉,等.我国苹果产业存在的主要问题、发展趋势及解决办法[J].江苏农业科学,2016,44(9):4-8.
[2]张铁中,杨丽,陈兵旗,等.农业机器人技术研究进展[J].中国科学:信息科学(增刊),2010(40):71-87.
[3]李占坤.果树采摘机器人控制系统研究与设计[D].镇江:江苏大学,2010.
[4]赵德安,沈甜,陈玉.苹果采摘机器人快速跟踪识别重叠果实[J].农业工程学报,2015,31(2):22-28.
[5]吕继东,赵德安,姬伟.苹果采摘机器人目标果实快速跟踪识别方法[J].农业机械学报,2014,45(1):65-72.
[6]司永胜,乔军,刘刚,等.苹果采摘机器人果实识别与定位方法[J].农业机械学报,2010,41(9):148-153.
[7]苑严伟,张小超,胡小安.苹果采摘路径规划最优化算法与仿真实现[J].农业工程学报,2009,25(4):141-144.
[8]TIEN THANH NGUYEN.Detection of red and bicoloured apples on tree with an RGB-D camera[J].Computers&Electronics in Agriculture,2009,66(1):85-92.
[9]张春龙,张楫,张俊雄,等.近色背景中树上绿色苹果识别方法[J].农业机械学报,2014,45(10):277-281.
[10]HU M K.Visual pattern recognition by moment invariants[J].IRETransactions on Information Theory,1962,8(2):179-187.
[11]李云,胡学龙.二值图像中标定目标区域的几何特征提取[J].计算机技术与发展,2000(5):55-57.
[12]任玉洁,付丽霞,张勇,等.基于平滑A*人工势场法的机器人动态路径规划[J].软件导刊,2018(1):26-29.
[13]张超超,房建东.基于定向加权A*算法的自主移动机器人路径规划[J].计算机应用,2017(S2):1245-1249.
[14]李挚,宋顶立,张双江,等.两种改进的最优路径规划算法[J].北京科技大学学报,2005,27(3):367-370.
[15]刘成菊,韩俊强,安康.基于改进RRT算法的RoboCup机器人动态路径规划[J].机器人,2017,39(1):86-88.
[16]陈春梅.无人机快速三维航迹规划算法的研究[D].成都:电子科技大学,2015.
[17]王鹏程,夏洁.基于特殊双向A*搜索算法的三维航迹规划[J].系统仿真学报:增刊,2009(20):229-233.
[18]邱磊.利用跳点搜索算法加速A*寻路[J].兰州理工大学学报,2015,41(3):102-107.
[19]祁悦,赵洋,杨帆.一种基于A*算法的分层路径规划在3D游戏中的应用研究[J].电子设计工程,2014,22(14):37-39.
[20]HARABOR D,BOTEA A.Breaking path symmetries on 4-connected grid maps[C].Proceedings of the 6th Artificial Intelligence and Interactive Digital Entertainment Conference,2010:33-38.
[21]石俊卫,包世泰,冯煜.基于A*算法的山地最优路径分析[J].现代计算机,2013(14):9-11.