Bag of feature approach for vehicle classification in heterogeneous traffic
Jyothy Das, Martish Shah, Leena Mary · 2017
In this paper, we are proposing Bag of Feature (BoF) approach for vehicle classification using Speeded Up Robust Features (SURF). First, monocular video taken using a stationary camera is given as the input to Gaussian Mixture Model (GMM) based foreground detector. Then a grid is used to measure the number of foreground pixels. If the pixels inside the grid is greater than a pre-assigned threshold, the grid will sense the presence of a vehicle. When the presence of a vehicle is identified within a frame, the region within the bounding box (created by foreground detector), which contain the RGB image of the vehicle is cropped and saved. SURF features of the stored frames are extracted and used for training classifier for 5 vehicle classes, namely, two wheelers, three wheelers, cars, Heavy Motor Vehicle (HMV) and Light Commercial Vehicle (LCV). For classifying a test video, frames with vehicles are extracted using the grid and SURF features are extracted and applied to the classifier. When the feature of test vehicle image becomes closer to that of class, the vehicle will be categorized into that class. Effectiveness of the proposed method is verified for different videos of heterogeneous traffic.