Abstract:
This paper addresses a framework for event decision of vision-based intelligent surveillance system based on the fuzzy model. The input probabilities of the tasks for the...Show MoreMetadata
Abstract:
This paper addresses a framework for event decision of vision-based intelligent surveillance system based on the fuzzy model. The input probabilities of the tasks for the fuzzy system are computed using the object detector which combines a cascade support vector machine (SVM) and color probability model (CPM). The SVM is used for identifying either the human, vehicle or baggage, while the CPM is applied to detect any possible smoke and fire regions in the monitoring area. The tracking algorithm is also integrated for triggering an alarm of suspicious event. The effectiveness of the proposed framework is evaluated under several video sequences with the comprehensive scenario. The results show that the framework can be one of the solutions for developing intelligent surveillance system.
Published in: 2017 IEEE International Conference on Mechatronics (ICM)
Date of Conference: 13-15 February 2017
Date Added to IEEE Xplore: 08 May 2017
ISBN Information: