High Performance Object Detection on Big Video Data Using GPUs

High resolution cameras have become inexpensive, compact and ubiquitously present in smart phones and surveillance systems. As a result huge volumes of images and video data are being generated daily. This availability of big video data has created challenges to video processing and analysis. Novel and scalable data management and processing frameworks are needed to meet the challenges posed by the big video data. This paper focuses on the first step in meeting this challenge that is to have high performance processing of big video data using GPUs. Parallel implementation of video object detection algorithm is presented along with fine grain optimization techniques and algorithm innovation. Experimental results show significant speedups of the algorithms resulting in real time processing of HD and beyond HD (like panoramic) resolution videos.

[1]  Ankush Mittal,et al.  Real-time moving object detection algorithm on high-resolution videos using GPUs , 2012, Journal of Real-Time Image Processing.

[2]  W. Eric L. Grimson,et al.  Adaptive background mixture models for real-time tracking , 1999, Proceedings. 1999 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (Cat. No PR00149).

[3]  R. Miyamoto,et al.  Parallel implementation of morphological processing on Cell/BE with OpenCV interface , 2008, 2008 3rd International Symposium on Communications, Control and Signal Processing.

[4]  Arie Shoshani,et al.  Optimizing connected component labeling algorithms , 2005, SPIE Medical Imaging.

[5]  Rama Chellappa,et al.  Object Detection, Tracking and Recognition for Multiple Smart Cameras , 2008, Proceedings of the IEEE.

[6]  Carl G. Looney,et al.  Fast connected component labeling algorithm using a divide and conquer technique , 2000, CATA.

[7]  Guna Seetharaman,et al.  Parallel Blob Extraction Using the Multi-core Cell Processor , 2009, ACIVS.

[8]  Sebastiano Battiato,et al.  Advanced Concepts for Intelligent Vision Systems , 2015, Lecture Notes in Computer Science.

[9]  Christophe Fiorio,et al.  Two Linear Time Union-Find Strategies for Image Processing , 1996, Theor. Comput. Sci..

[10]  Ankush Mittal,et al.  OS-Guard: on-site signature based framework for multimedia surveillance data management , 2010, Multimedia Tools and Applications.

[11]  Ian D. Reid,et al.  Robust Real-Time Visual Tracking Using Pixel-Wise Posteriors , 2008, ECCV.

[12]  Chun-Jen Chen,et al.  A linear-time component-labeling algorithm using contour tracing technique , 2004, Comput. Vis. Image Underst..

[13]  Trista Pei-chun Chen,et al.  Computer Vision Workload Analysis: Case Study of Video Surveillance Systems , 2005 .

[14]  Kentaro Toyama,et al.  Wallflower: principles and practice of background maintenance , 1999, Proceedings of the Seventh IEEE International Conference on Computer Vision.

[15]  Zoran Zivkovic,et al.  Improved adaptive Gaussian mixture model for background subtraction , 2004, Proceedings of the 17th International Conference on Pattern Recognition, 2004. ICPR 2004..