GLCM based analysis of combustion flame parameters using edge detection of flame radiation images

Monitoring and analysis of industrial flames is becoming progressively important to achieve comprehensive understanding and successive optimization of combustion process. This is becoming increasingly a prominent research field in order to increase combustion efficiency and to decrease toxic pollutant emissions. Now, in recent years, with the arrival of digital image processing techniques, CCD (Charge Coupled Device) or vision based analysis, the characterisation of combustion flames have been developed. This paper presents GLCM (Gray Level Co-occurance Matrix) based two-dimensional (2D) imaging technique to measure various geometric & luminous parameters from the radiation images of a pulverized coal flame. LBP (Local binary Pattern) based edge detection of red color component of radiation images at different temperatures is carried out. With the help of GLCM of radiation gray scale images, various statistical image features are calculated & correlated with certain flame parameters. Experimental results provide a visual inspection to predict flame parameters variation with respect to the temperature through the analysis of statistical features of the image.