基于机器人仿生视觉路标多模式匹配定位仿真
|
摘要
针对传统机器人仿生视觉路标多模式匹配定位过程中,普遍存在着定位运行时间较长、准确度较低、成本消耗较大等问题。提出一种新型机器人仿生视觉路标多模式匹配定位方法。通过对机器人仿生视觉路标进行分析,同时考虑到机器人定位时的实际方向与视觉路标指示的方向存在一定的偏差,对方向偏差进一步修正,得到当前位置较为准确的机器人姿态信息;采用双目视觉传感器采集路标图像中的匹配点对,获取视觉路标图像三维坐标,构建双目立体视觉定位模型,根据该模型获得机器人与路标的相对位置关系,实现机器人视觉路标多模式匹配定位。实验结果表明,所提方法定位运行时间较短、准确度较高、成本消耗较低。
Due to long running time, low accuracy and high cost in the traditional multi-pattern matching and positioning of robot bionic visual guidepost, this paper puts forward a new method of multi-pattern matching and positioning for robot bionic visual guidepost. Based on the analysis of robot bionic visual guidepost and the consideration for the deviation between the actual direction of the robot and the direction indicated by visual guidepost, the deviation was further corrected to obtain the more accurate attitude information of robot at the current position. After that, the binocular vision sensor was used to collect the matching pairs in guidepost image, so as to obtain the three-dimensional coordinate of visual guidepost image. Finally, the binocular stereo vision positioning model was built. According to this model, the relative position relationship between robot and guidepost was obtained. Thus, the multi-pattern matching positioning of robot vision guidepost was achieved. Simulation results show that the proposed method has shorter operation time, higher accuracy and lower cost during the location.
Due to long running time, low accuracy and high cost in the traditional multi-pattern matching and positioning of robot bionic visual guidepost, this paper puts forward a new method of multi-pattern matching and positioning for robot bionic visual guidepost. Based on the analysis of robot bionic visual guidepost and the consideration for the deviation between the actual direction of the robot and the direction indicated by visual guidepost, the deviation was further corrected to obtain the more accurate attitude information of robot at the current position. After that, the binocular vision sensor was used to collect the matching pairs in guidepost image, so as to obtain the three-dimensional coordinate of visual guidepost image. Finally, the binocular stereo vision positioning model was built. According to this model, the relative position relationship between robot and guidepost was obtained. Thus, the multi-pattern matching positioning of robot vision guidepost was achieved. Simulation results show that the proposed method has shorter operation time, higher accuracy and lower cost during the location.
引文
[1] 王映知,周翕,王永,等.并联机器人运动目标精确优化定位仿真研究[J].计算机仿真,2017,34(10):320-324.
[2] 张智,张磊,苏丽,等.基于人工离线特征库的室内机器人双目定位[J].哈尔滨工程大学学报,2017,38(12):1906-1914.
[3] 李珣,刘丽,洪良,等.移动机器人室内无源RFID定位方法及实现[J].计算机工程与应用,2017,53(16):230-236.
[4] 张涛,马磊,梅玲玉.基于单目视觉的仓储物流机器人定位方法[J].计算机应用,2017,37(9):2491-2495.
[5] 陈英.工业机器人交流伺服驱动系统研究[J].电子设计工程,2017,25(16):80-83.
[6] 冯晟,吴成东,张云洲.基于改进APIT的移动机器人动态定位[J].北京邮电大学学报,2016(5):178-182.
[7] 温熙,郭杭.室内移动机器人自定位方法[J].测绘科学,2016,41(6):97-101.
[8] 李朕阳,郎朗,陈孟元.基于SR-CKF的相对方位多机器人协同定位算法[J].电子测量与仪器学报,2016,30(7):1107-1113.
[9] 任红格,刘伟民,李福进.一种记忆可修剪型仿生机器人的速度跟踪算法研究[J].现代电子技术,2017,40(15):149-153.
[10] 林辉灿,吕强,王国胜,等.基于VSLAM的自主移动机器人三维同时定位与地图构建[J].计算机应用,2017,37(10):2884-2887.
[11] 陈晓飞,凌有铸,陈孟元.WSNs环境下基于高斯混合容积卡尔曼滤波的移动机器人定位算法?[J].传感技术学报,2017,30(1):133-138.
[12] 黄显良,郁建芳,戚浩,等.安徽霍山窗中小地震活动与精定位研究[J].地震工程学报,2016,38(2):236-241.
[2] 张智,张磊,苏丽,等.基于人工离线特征库的室内机器人双目定位[J].哈尔滨工程大学学报,2017,38(12):1906-1914.
[3] 李珣,刘丽,洪良,等.移动机器人室内无源RFID定位方法及实现[J].计算机工程与应用,2017,53(16):230-236.
[4] 张涛,马磊,梅玲玉.基于单目视觉的仓储物流机器人定位方法[J].计算机应用,2017,37(9):2491-2495.
[5] 陈英.工业机器人交流伺服驱动系统研究[J].电子设计工程,2017,25(16):80-83.
[6] 冯晟,吴成东,张云洲.基于改进APIT的移动机器人动态定位[J].北京邮电大学学报,2016(5):178-182.
[7] 温熙,郭杭.室内移动机器人自定位方法[J].测绘科学,2016,41(6):97-101.
[8] 李朕阳,郎朗,陈孟元.基于SR-CKF的相对方位多机器人协同定位算法[J].电子测量与仪器学报,2016,30(7):1107-1113.
[9] 任红格,刘伟民,李福进.一种记忆可修剪型仿生机器人的速度跟踪算法研究[J].现代电子技术,2017,40(15):149-153.
[10] 林辉灿,吕强,王国胜,等.基于VSLAM的自主移动机器人三维同时定位与地图构建[J].计算机应用,2017,37(10):2884-2887.
[11] 陈晓飞,凌有铸,陈孟元.WSNs环境下基于高斯混合容积卡尔曼滤波的移动机器人定位算法?[J].传感技术学报,2017,30(1):133-138.
[12] 黄显良,郁建芳,戚浩,等.安徽霍山窗中小地震活动与精定位研究[J].地震工程学报,2016,38(2):236-241.