Vehicles Detection System at Different Weather Conditions

The importance of efficient vehicle detection (VD) is increased with the expansion road networks and number vehicles in Intelligent Transportation Systems (ITS). This paper proposes a system for detecting at different weather conditions such as sunny, rainy, cloudy foggy days. first step to proposed implementation determine whether video’s condition normal or abnormal. Random Forest (RF) classification was performed video while features were extracted two frames by using Gray Level Co-occurrence Matrix (GLCM). In this system, background subtraction applied mixture Gaussian 2 (MOG 2) then applying pre-processing operations, blur filter, dilation, erosion, threshold. main contribution propose histogram equalization technique complex instead filter improving clarity, which leads increase accuracy. Based on previous steps, defines each region frame expected contain vehicles. Finally, Support Vector Machine (SVM) classifies defined regions into not. As compared methods, showed high efficiency about 96.4% ability detect conditions.