CAJ | 학술논문

介绍了一种基于颜色分割和区域描述的交通标志检测方法.该方法利用在RGB空间的颜色聚类算法分割出感兴趣色彩区域,然后进行形态学处理,最后结合Hu不变矩对道路交通标志进行识别.识别结果表明,基于颜色聚类和Hu不变矩的交通标志识别方法具有很强的抗图像平移、缩放和旋转识别能力,并具有实现简单、识别速度快、准确率高等特点,有较高的实用价值.
개소료일충기우안색분할화구역묘술적교통표지검측방법.해방법이용재RGB공간적안색취류산법분할출감흥취색채구역,연후진행형태학처리,최후결합Hu불변구대도로교통표지진행식별.식별결과표명,기우안색취류화Hu불변구적교통표지식별방법구유흔강적항도상평이、축방화선전식별능력,병구유실현간단、식별속도쾌、준학솔고등특점,유교고적실용개치.
@@@@Abstact:A method of road traffic sign detection based on color segmentation and region descriptions is introduced. The method firstly uses color clustering algorithm in RGB space to segment the interested color area, and then conducts an morphological processing, and Finally recognize the traffic sign by using the Hu invariant moments. The results show that:the traffic sign recognition method based on color clustering and Hu invariant moments has a strong robustness to image translation, scaling and rotation, and this method has the features of simply implement, fast recognition and high accuracy, it has high practical value.