A comprehensive approach for road marking detection and recognition

• Published in: Multimedia tools and applications Vol. 79; no. 23-24; pp. 17193 - 17210
• Main Authors: Ding, Ling, Zhang, Huyin, Xiao, Jinsheng, Li, Bijun, Lu, Shejie, Klette, Reinhard, Norouzifard, Mohammad, Xu, Fang
• Format: Journal Article
• Language: English
• Published: New York Springer US 01.06.2020; Springer Nature B.V
• Subjects: Algorithms; Computer Communication Networks; Computer Science; Data Structures and Information Theory; Image detection; Marking; Multimedia Information Systems; Object recognition; Special Purpose and Application-Based Systems; Support vector machines; Road markings; Image segmentation; PCA-HOG; SVM; Object detection
• ISSN: 1380-7501; 1573-7721
• DOI: 10.1007/s11042-019-08384-5

This paper presents a comprehensive approach for road marking detection and recognition which collect the image information from a camera mounted in the front of the vehicle. This approach uses PCA-HOG feature and support vector machine (SVM) to detect and recognize the road markings. First, the vanishing point in the image is identified, and is used to generate a bird’s-eye view of the road. Second, the method of local median binarization is used to segment the image and it can avoid the interference of light and shadow. Finally, the feature vector is generated by using a combination of HOG and PCA, and the SVM is used to determine the feature vector as a specific road marking. The test results show that this algorithm higher accuracy and recall rate. Moreover, the algorithm has low computational complexity and can guarantee high recognition speed. Therefore, compared with existing algorithms, it has great advantages. This research will play an important role and a wide range of applications in advanced driver assistant system (ADAS) or driverless driving.