The Fire Alarm And Evacuation Strategy Of Ship Fire Disaster Based On Vision Sensors

Fire disaster does great danger to the safety of ships. Since the 80 s of last century, the threats of marine fire disasters have been paid a lot of attentions. Researches on the fire disasters generally focus on three aspects: early detection of the fire, the process of ship fire simulation, and the research on the fire escape. From the research literature of the past, researchers tend to separate the three aspects. This article studies from a new Angle, combine these areas, realize the innovations in the following:By using the theory of thermodynamics of smoke movement of diffusion and heat transfer analysis, the conclusion of the difference of smoke movement and rigid body movement is reached. Then we use image entropy to detect smoke from images; Use Star-ccm to simulate the smoke movement in local area in a ship. After dealing with the result, we combine the simulation and the observation that is processed by digital image processing together. Then we can compare their common characters; To realize escape personnel identification, through using template matching method and string matching method, to match and analysis the picture frame differential which is captured by visual sensors, and get the information of the amount of escaping people, and then navigate the escape personnel.The theory analysis part analyzes the movement of the smoke. smoke movement has the same rule of heat transfer theory and fits the characteristics of free jet flow movement in the actual movement. On the information entropy, we point out that its similarity in essence and thermodynamic entropy. We have carried on the dynamic analysis model, points out main factors affecting the efficiency of the escape.Experimental section of smoke recognition method based on image entropy is verified, the results in the frequency spectrum analysis are ideal.Digital image processing part has implemented the smoke areas in image segmentation successfully. Then we match the results of simulation with the observed results. To realize the number identification and match the theoretical calculation of the channel’s largest survival ability, we achieve the optimal allocation of the channel number to escape.