基于小波包的机器人多目标识别与定位研究  被引量:1

Research on Robot Multi-Target Recognition and Localization Based on Wavelet Packet

在线阅读下载全文

作  者:徐亭婷 段锁林[1] 庄玮[1] 

机构地区:[1]常州大学机器人研究所,江苏常州213164

出  处:《计算机仿真》2014年第4期374-378,共5页Computer Simulation

摘  要:针对GRB-400视觉机器人系统,研究了一种基于小波包分解的图像匹配算法并将其用于机器人多目标识别与定位系统。上述系统利用基于小波包分解的算法对获取到的多目标图像的高频部分进行处理,去除了高频噪声干扰的同时获得了目标物体的高效边缘信息。然后运用Hu矩不变量构成目标图像的边缘矩特征向量,再利用欧几里得距离来度量各个物体与模板之间的相似度以完成图像匹配,最终在多目标中获得了待抓取的目标物体。随后利用矩质心法获得待抓取目标的质心位置,实现了目标物体的定位。仿真结果表明,上述算法在保证运行速度的同时,能有效地完成机器人对多目标物体的识别及定位。According to the GRB - 400 robot system, a novel method of image matching algorithm based on wave let package was adopted in the robot system for the multi target recognition and localization. First, in order to ob tain the hight efficient edge information of target objects while removing the high frequency noise in the image, the o riginal image was decomposed to two levels by wavelet packet. Only the high frequency part of the image was pro cessed during the whole processing. Then Hu moment invariants were applied to measure the similarity between ob jects and template to complete image matching, and ultimately to obtain the target object for robot grabbing. Finally, the movement qualitative method was used to acquire the center of mass position of the target object, so as to com plete later target object grabbing. The simulation results show that the adopted method can help the robot system ef fectively complete multi objective object recognition and localization, at the same time ensure the runnin sneed.

关 键 词:机器人 小波包 多目标识别 定位 

分 类 号:TP242[自动化与计算机技术—检测技术与自动化装置]

 

参考文献:

正在载入数据...

 

二级参考文献:

正在载入数据...

 

耦合文献:

正在载入数据...

 

引证文献:

正在载入数据...

 

二级引证文献:

正在载入数据...

 

同被引文献:

正在载入数据...

 

相关期刊文献:

正在载入数据...

相关的主题
相关的作者对象
相关的机构对象