Vehicle Counting using Deep Learning Models: A Comparative Study

Recently, there has been a shift to deep learning architectures for better application in vehicle traffic control systems. One popular deep learning library used for detecting vehicle is TensorFlow. In TensorFlow, the pre-trained model is very efficient and can be transferred easily to solve other similar problems. However, due to inconsistency between the original dataset used in the pre-trained model and the target dataset for testing, this can lead to low-accuracy detection and hinder vehicle counting performance. One major obstacle in retraining deep learning architectures is that the network requires a large corpus training dataset to secure good results. Therefore, we propose to perform data annotation and transfer learning from an existing model to construct a new model for vehicle detection and counting in the real world urban traffic scenes. Then, the new model is compared with the experimental data to verify the validity of the new model. Besides, this paper reports some experimental results, comprising a set of innovative tests to identify the best detection algorithm and system performance. Furthermore, a simple vehicle tracking method is proposed to aid the vehicle counting process in challenging illumination and traffic conditions. The results showed a significant improvement of the proposed system with the average vehicle counting of 80.90%.