Abstract
Road traffic automatic incident detection system is the important composition of a road traffic monitoring system. On the basis of support vector machines (SVM) theory and sequential minimal optimization (SMO) algorithm, the road traffic incident detection model based on SMO was put forward. With Visual C++ and Matlab, a simulation experiment was applied to the model. The shortcomings of quadratic programming (QP) algorithm were analyzed. The effects of three different kernel functions on sample training SVM classifier and on detection performance were compared using Chunking algorithm and SMO algorithm. The results show that Gauss kernel function is better than the other two kernel functions for the performance of training and incident detection using SMO.
Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
