A large number of security-related problems could be addressed by night time video surveillance. This has additional challenges of handling nonuniformly illuminated areas in the frame as compared to day time videos. The presence of vehicle lights degrades the existing background extraction process. An intensity- based shot boundary detection technique is applied for the night time videos in this paper. This algorithm is used to identify the abrupt transition frames that are caused by the vehicle lights or moving lights. The frames occuring between such detected frames could be reliably used for background extraction and night time video enhancement. The algorithm is tested on few real-time videos and their results are provided. © Springer India 2016.
cited By 0; Conference of International Conference on Soft Computing Systems, ICSCS 2015 ; Conference Date: 20 April 2015 Through 21 April 2015; Conference Code:160689
Dr. Padmavathi S. and Abirami, G., “Detection of abrupt transitions in night time video for illumination enhancement”, Advances in Intelligent Systems and Computing, vol. 397, pp. 809-817, 2016.