Visible to the public Tracking and Abnormal Behavior Detection in Video Surveillance Using Optical Flow and Neural Networks

TitleTracking and Abnormal Behavior Detection in Video Surveillance Using Optical Flow and Neural Networks
Publication TypeConference Paper
Year of Publication2014
AuthorsRasheed, N., Khan, S.A., Khalid, A.
Conference NameAdvanced Information Networking and Applications Workshops (WAINA), 2014 28th International Conference on
Date PublishedMay
Keywordsabnormal behavior detection, Adaptive optics, chaotic movement, Computer vision, feature selection, feed forward neural network, feedforward neural nets, FGMM model, Foreground Detection, foreground detection with Gaussian mixture model, Gaussian Mixture Models, Gaussian processes, image motion analysis, image sequences, Lucas-Kanade approach, mixture models, Neural Network, Neural networks, normal movement, object detection, Optical computing, optical flow, Optical imaging, real time videos, Streaming media, synthesized videos, targets identification, video frames, video surveillance

An abnormal behavior detection algorithm for surveillance is required to correctly identify the targets as being in a normal or chaotic movement. A model is developed here for this purpose. The uniqueness of this algorithm is the use of foreground detection with Gaussian mixture (FGMM) model before passing the video frames to optical flow model using Lucas-Kanade approach. Information of horizontal and vertical displacements and directions associated with each pixel for object of interest is extracted. These features are then fed to feed forward neural network for classification and simulation. The study is being conducted on the real time videos and some synthesized videos. Accuracy of method has been calculated by using the performance parameters for Neural Networks. In comparison of plain optical flow with this model, improved results have been obtained without noise. Classes are correctly identified with an overall performance equal to 3.4e-02 with & error percentage of 2.5.

Citation Key6844614