摘要
提出了一种基于交警手势识别的交通灯控制方法.该方法首先对彩色摄影机采集的交警手势图像进行预处理,包括彩色图像平滑、彩色图像增强;其次利用边缘检测和图像分割方法对目标图像进行分割,由于图像受噪声干扰严重,采用Sobel算子消除噪声;再次将分割得到的目标轮廓进行表示与描述,为图像的识别作好准备;最后结合图像的轮廓特征对图像进行识别,采用了模糊模式识别方法.试验结果表明,该方法能够实现对交警手势准确的识别.
Create a way to control the traffic light which is achieved by identifying the pointsman＇ s gestures. In this way, at the beginning, pretreating the image of pointsman＇s gesture from the eolour vidicon which includes the colour image smoothing and intensifying. Then using the way of edge checking and image segmentation to divide the target image, due to the image is interfered badly by noise, Sobel operator is used. Presenting and discribing the divided target profile to prepare the image identifying. Finally, identifying the image with the help of the picture＇ profile characteristic. Fuzzy model is taken in this paper, and experiment result shows that this way is able to identify the points- man＇s gesture correctly.
出处
《北京工商大学学报（自然科学版）》
CAS
2009年第1期33-36,共4页
Journal of Beijing Technology and Business University:Natural Science Edition
关键词
图像处理
图像描述
图像识别
image procession
image description
image identification
