Vehicle detection using histogram of oriented gradients and real adaboost

The vehicle detection system is an important technology because it has many applications in traffic fields such as traffic monitoring, counting the number of vehicles passing, calculating the speed of an oncoming vehicle, and so on. Histogram Of Oriented Gradients (HOG) is a feature descriptor used for object detection. HOG describes features based on some local histogram of gradient orientation weighted by gradient magnitude. Real AdaBoost is a learning algorithm which combined weak classifier into a strong classifier that represents the final output of the boosted classifier. This research aims to detect vehicles in the static image using a Histogram Of Oriented Gradients methods and Real Adaboost. The steps of vehicle detection are pra-processing, feature extraction process with HOG, and classification process with Real AdaBoost. The testing result shows that the system can detect vehicles with an accuracy level of 91.78 % from 292 testing images, from 257 images of the vehicle and 35 images of not vehicle.