摘要
针对机器人在电子地图中的定位问题,基于二进制图案编码方式,提出了编码区域的定位算法和编码图案的识别算法。该算法利用编码图案与地面之间的差别自适应选择阈值,实现编码图案的分割;使用直线扫描法确定编码图案的顶点信息,定位编码区域;运用先验结构信息对分割出的编码区域进行识别。另外,分析了影响算法精度的环境光照、噪声等因素,提出相应的解决方案。实验表明该算法实现简单、速度快、定位和识别精度高,准确率达到97%。
As of the problem of robot localization ,we put forward an algorithm for the localization of encoding region and the recog nition of coding pattern by use of the binary pattern encoding method. This method chooses the adaptive threshold according to tile difference between the encoding pattern and the ground and realizes encoding pattern segmentation, determines the vertices of encoding pattern by using the line scanning algorithm and locates the encoding area, and recognizes the segmented encoding patterns by using the priori knowledge of the encoding area. The influence of such factors as environmental illumination, noise, et al, upon the accuracy of the algorithm is analyzed, and the corresponding solutions are suggested. Experiments show that the algorithm is quick and easy to be imple mented with high accuracy of orientation and recognition, and the accuracy is up to 97 %.
出处
《济南大学学报（自然科学版）》
CAS
北大核心
2014年第3期184-188,共5页
Journal of University of Jinan(Science and Technology)
基金
国家自然科学基金青年基金(61203341)
山东省高等学校科技计划(J12LN19)
关键词
机器人
电子地图
定位
二进制编码
直线扫描
robot
electronic map
localization
binary encoding pattern
line scanning
