Real-Time Dynamic Scene Analysis Using Differential Technique and Performance Evaluation with Optical Flow

Real time dynamic scene analysis has become very important aspect as the increase in video input analysis. Although several dynamic scene analysis techniques are available, some of them poses increased computational complexity problem. In the present work a simple differential algorithm is designed and tested with traffic flux estimation application. Traffic flux estimation will play a very vital role in implementing intelligent traffic control scheme. The technique developed is having simple statistical background. Dynamic selection of images, from the sequence is implemented successfully in order to reduce the computation time. To establish the validity of these techniques, many natural traffic sequences have been obtained without considering any preconceived conditions like illumination, obstacles and shadows. While, the sequences selected are devoid of ego-motion problem. The designed techniques are evaluated with such twenty different video sequences and weighed thoroughly with simple confidence measures. In the present work we have achieved real time analysis with normal video rate of 30 frames per second. This is in comparison with the one frame per second using near real time optical flow computation. The result produced with this analysis is extremely good and is beneficial in the real time traffic control and tracking of vehicles in the urban areas. MATLAB image processing toolbox is explored to implement the techniques.

[1]  Joseph K. Kearney,et al.  Optical Flow Estimation: An Error Analysis of Gradient-Based Methods with Local Optimization , 1987, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[2]  Yiannis Aloimonos,et al.  The Statistics of Optical Flow , 2001, Comput. Vis. Image Underst..

[3]  Brendan McCane,et al.  On Benchmarking Optical Flow , 2001, Comput. Vis. Image Underst..

[4]  Rama Chellappa,et al.  Accuracy vs Efficiency Trade-offs in Optical Flow Algorithms , 1996, Comput. Vis. Image Underst..

[5]  Tomaso A. Poggio,et al.  Motion Field and Optical Flow: Qualitative Properties , 1989, IEEE Trans. Pattern Anal. Mach. Intell..

[6]  Berthold K. P. Horn,et al.  Determining Optical Flow , 1981, Other Conferences.

[7]  Keith Langley,et al.  Recursive Filters for Optical Flow , 1995, IEEE Trans. Pattern Anal. Mach. Intell..

[8]  David J. Fleet,et al.  Performance of optical flow techniques , 1994, International Journal of Computer Vision.

[9]  Martin Herman,et al.  Real-time obstacle avoidance using central flow divergence and peripheral flow , 2017, Proceedings of IEEE International Conference on Computer Vision.