Performance Enhancement of Multi-Camera Handoff Scheme using ANFIS Method

The multicamera system aims to monitor a human target for surveillance activity in wide-area observation. The monitoring task will be done by the cameras using a handover process among them, following the moving target. The most important thing for the handover scheme is the execution time to do the decision-making process by the server. The slower the execution time of the handover decision, the slower the target is handled by other cameras. The consequence of this case is there is the possibility that the target will escape from the monitoring system. This paper provides a development technique for hand-offering human targets in the multicamera system based on the fuzzy logic decision-making process. An Adaptive Neuro-Fuzzy Inference System (ANFIS) method is proposed as a decision-making process to decide whether a human target should be handed over from one camera to another camera or not. The performance of the proposed method was examined by comparing it with the Cascaded-FIS method. The simulation result shows that the computation time to make the handover process has been reduced up to 62.88 % by utilizing the proposed method, faster than the Cascaded-FIS method.