A Study of Traffic Line Detection using Hough Neural Network

• Main Authors: Wei-Jen chen, 陳韋任
• Other Authors: Yih-Lon Lin
• Format: Others
• Language: zh-TW
• Published: 2011
• Online Access: http://ndltd.ncl.edu.tw/handle/42393065838944391428

碩士 === 義守大學 === 資訊工程學系碩士班 === 99 === In image processing, detected shapes are very important, which are often in the forms of lines or circulars. Detected edges usually exhibit broken or discontinuous features. Hough transform is the algorithm that can links the edges together when it is of broken or discontinuous states. This study applies Hough transform using neural network to improve the accuracy of edge detections. For traffic images, results of line detection are effected by environmental factors or other noise such as impact of light and weather changes. By using Hough neural network, we increase the accuracy of line detection to get lane lines on traffic images.