Research On Road Traffic Sign Detection And Recognition Technology

Traffic sign recognition system is an important part of intelligent transportation systemsand advanced driver assistance systems to improve the accuracy of the traffic sign detectionand recognition algorithm and real-time towards the key issues need to be addressed in theactual application process. The accuracy of the algorithm is a very important factor in thestudy of traffic sign recognition, wrong recognition result not only can not play the role ofadjuvant driving, but also lead to serious security incidents. The real-time nature of thealgorithm determines the research results can be translated into practical value products. Thenumber of cars increasing traffic accidents remains high, rising car driving intelligent realitythe face of pressure, to carry out real-time applications that target traffic sign detection andrecognition technology, is of great significance to increase the driving safety.In this paper, traffic sign detection and recognition accuracy and real-time issues as themain object of study, from image preprocessing traffic signs, detection and segmentation,feature extraction and recognition aspects of this issue extensive and in-depth research. Theinnovation of this paper as well as the main duties includes:1. Traffic sign image pre-processing technology used to improve the quality of theacquired image histogram equalization and Gaussian filter-based approach, while ensuringthat the pretreatment of computing time in the affordable range;2. Under any scenario studied traffic sign detection and segmentation techniques, makefull use of traffic signs and geometric shapes color threshold characteristic parameters toachieve a variety of complex scenarios of accurate positioning and segmentation, a largenumber of experiments show that the algorithm for the day, night, rain traffic signs and otherscenarios have better positioning effect.3. Studied based on the traffic sign recognition template matching technique is proposedbased on SURF traffic sign recognition algorithm, feature extraction algorithm uses SURFoperator to describe the characteristics of traffic signs, and finally through the establishmentof the library search feature template judged by comparing the traffic to be recognized type ofsign, experiments show that the algorithm’s strategy with good results.4. Studied based on Zernike moments and support vector machine traffic signrecognition technology for the template matching method based on SURF scalability is poor,do not have the self-learning ability and other issues, the use of Zernike moment invariants to build a collection of traffic signs feature vectors, using the support vector machine to achievefast and accurate classification, through comparison with other methods, experimental resultsshow that this method has higher accuracy and good real-time.Finally, the full text of the summary and outlook traffic sign detection and recognitionrequires further study and practical application prospects.