Calculation Methods Of Hearth Combusting Flame's Temperature Based On Image Processing

Modern industry expects a real time and high precision temperature distribution test method, which would be positively significant in improving the quality of products, optimizing burning process, saving energy resource, protecting environment and production safety. The flame of hearth is the most direct and quick expression of status of the hearth. So, it is a new technology and still a task of great challenge to use color CCD camera to acquire flame image and calculate the temperature that based on image processing.In order to calculate the temperature of hearth's combusting flame, with the image processing technology and the stakeout equipments on hearth's combusting flame, the region of interesting flame image will be extracted based on the thermal radiation theory and the foundation of color CCD camera. With the designed procedure and the simulated software, the temperature of hearth's combustion flame is analyzed. The main job includes the following aspects:Firstly, in order to realize the temperature measurement on static state flame image, the thesis introduces several kinds of means including colorimetric temperature-measurement methods and its improved measurement, the least square method and BP(Back-Propagation) neural network method, an improved method ofBP neural network based on HSI image mode and another method based on colorimetric temperature- measurement, the latter is more practical and precise.Secondly, with proper measurement, the average temperature and the relation curve about temperature and time is calculated to view the whole combustion state.Finally, after cropping the dynamic flame image and studying the relation between the gray-level image intensity and temperature, the simulated software show the average temperature of dynamic cutting area and the real time curve about temperature and time, which has been used to view the dynamic combustion state.The results of experiments show that the algorithm is effective, and the system is high effective and reliable.