Vehicle detector training with labels derived from background subtraction algorithms in video surveillance

摘要

Vehicle detection in video from a miniature stationary closed-circuit television (CCTV) camera is discussed in the paper. The camera provides one of components of the intelligent road sign developed in the project concerning the traffic control with the use of autonomous devices being developed. Modern Convolutional Neural Network (CNN) based detectors need big data input, usually demanding their manual labeling. In the presented research approach the weakly-supervised learning paradigm is used for the training of a CNN based detector employing labels obtained automatically through an application of video background subtraction algorithm. The proposed method is evaluated on GRAM-RTM dataset and a CNN fine-tuned with labels from the background subtraction algorithm. Even though obtained representation in the form of labels may include many false positives and negatives, a reliable vehicle detector was trained employing them. The results are presented showing that such a method can be applied to traffic surveillance systems.