机器人视觉系统模糊识别抓取物算法.PDFVIP

2009 2 F ebru ary 2009 35 2 Journa l o f Be ijing U nivers ity of A eronautics and A stronautics V o.l 35 N o1 2 黄荣瑛 ( , 10019 1) 针对机器人视觉系统对抓取物的模糊识别问题, 参照人眼-脑识别对象的过 程, 建立 包括区域分割与模糊识别两个环节的识别算法. 以视觉图像的灰度与色度特征作为 区域分割与模糊识别的依据, 从灰度色度及形体特征上提取特征集的指标, 并根据经 w a lsh 变换后图像灰度迅速向低频聚集的特点, 提出基于 w alsh 变换的基元模式识别特征的定义方 法. 在构建识别目标矩阵与关系矩阵的基础上, 应用模糊关系合成与最大隶属度原则建立识别 算法. 该算法可从少量的采样点中识别出对象, 具有较好的实时性. 视觉; 图像分析; 模式识别; w a lsh 变换; 内积 TP 391 A 1001-5965 ( 2009) 02-0197-04 Fuzzy recogn ition and ob jec-t capture a lgo rithm fo r robot v is ion sy stem s H u ang R ongy ing ( Schoo l ofM echan ical Eng ineering and A u tom at ion, Beij ing U n iv ersity of A eron aut ics and A stronau tics, B eijing 100 191, Ch ina) Abstract W ith reference to the object recognition process of hum an eyes and m inds, the reg ion segm en- tation and fuzzy recognition a lgorithm s w ere estab lished fo r robot v ision system s in order to dea lw ith the prob- lem o f the object-capture recogn ition. The gray-sca le and chrom aticity o f mi ages w ere taken as a b asis for re- gion segm en tation and fuzzy recogn ition, the index o f the ch aracteristicsw ere taken from g ray- sca le, chrom atic-i ty and body features, and the definition m ethod fo r feature recogn it ion on the b asis o f w alsh tran sform m eta- m ode lsw as put forw ard w ith the sh arp low-frequ ency gathering tendency of the g ray- sca le. Then, the recogn-i t ion algorithm sw asw orked out ut iliz ing the fuzzy re lation com position and m ax mi um sub jection accord ing to the ob ject recogn it ion m atrix and relat ion m atrix w as estab