A Real-time Recognition System for Circular Signal Based On Hough Transform And Parallel Computing
Ship traffic lights is a constantly changing light and dark circular spot. To quickly and error-free receive the signal light in an complex environment the recognition system needs to identify intensity and shape of the light signal. Hough transform is a description for the boundary and shape of area and commonly be used to extract curve from a complex background. But its two disadvantages, massive storage requirement and high cost of computation, limited its application in real-time recognition system. This paper presents an integrated optimization algorithm which is focus on reducing the sampling points, reducing dimensions and utilize parallel computing methods. The system can recognize circular signal in real-time from a series of continuous images(size: 800X600 pixels) and make close to 100％ accuracy. At the same time as the characteristics of parallel computing, the computing power of the system is easy to expand.
hough transform circle recognition parallel compute realtime detection
Cui zhi Ge youhua Li yongjian Fang yong
Dept. of Mechanical Engineering Yancheng Institute of Technology Yancheng, China Patent Examination Cooperation Center Of SIPO Beijing, China
国际会议
重庆
英文
189-192
2011-01-21（万方平台首次上网日期，不代表论文的发表时间）