基于SSA优化SVM的交通标线识别算法研究

Research on Traffic Marking Recognition AlgorithmBased on SSA Optimized SVM

针对智能驾驶汽车对交通标线识别率不高的问题,提出一种改进LBP特征和SSA优化SVM的交通标线识别算法.该算法采用改进的LBP算法提取路面交通标线特征,通过PCA降维处理,获得最佳维度;通过对SSA、PSO、WOA和GWO 4种群智算法分析比较,发现采用SSA优化SVM的惩罚系数和核函数参数,获得路面交通标线识别的最优模型.实验结果表明,该模型的平均识别速度分别比PSO、WOA和GWO优化的SVM模型提高了1.0%、6.19%和8.35%,平均识别精度最高,达到96.7%,比优化前提高了4.5%,且模型的最佳参数稳定.

Aiming at the problem that the recognition rate of intelligent driving vehicles for traffic markings is not high,a traffic markings recognition algorithm based on improved LBP feature and SSA optimized SVM is proposed.The improved LBP algorithm is used to extract the characteristics of road traffic markings,and the best dimension is obtained by PCA dimension reduction.Through the analysis and comparison of SSA,PSO,WOA and GWO intelligent algorithms,SSA is used to optimize the penalty coefficient and kernel function parameters of SVM,and the optimal model of road traffic marking recognition is obtained.The experimental results show that the average recognition speed of the model is 1.0%,6.19%and 8.35%higher than that of the PSO,WOA and GWO optimized SVM models respectively,the average recognition accuracy is the highest at 96.7%,which is 4.5%higher than that before the optimization,and the optimal parameters of the model are the most stable.