Realtime fire detection using CNN and search space navigation

Intelligent search techniques and an intelligent agent for smart search are useful in many application domains. We develop a state space navigational model for intelligent agents aimed at industrial surveillance from fire hazards. Our focus is on fire detection using the convolution neural network then proactively search the area which is more likely to have routes toward the target. This problem can be simulated into an optimization problem over a state space, which can be figure out effectively through a greedy algorithm. We also compare our approach with both uninformed and informed search algorithms. We evaluate our proposed system using various search algorithms for search and rescue agent. The analysis of the results obtained demonstrate the efficiency of the system.