The present goal of combustion research is to enhance the burning efficiency resulting in minimal emission which is in fact, paves the way for a sustainable future. Researchers are investigating different parameters and factors associated with combustion to control the combustion process. Image processing is one of the most useful and safe tool for this job as it is nonintrusive and do not interfere with the combustion zone during experiment. Present work focuses on the digital image processing of the premixed and diffusion flame which has been utilized as a tool to characterize burner stabilized premixed and non-premixed Flame. The experiment is performed on a burner stabilized LPG-air flame. For premixed flame, several sets of experiments are done keeping the camera setting and image quality identical which resulted in an almost linearly increasing average RGB value with respect to equivalence ratio. Taking the relation of an experiment as standard, equivalence ratio is calculated for other experiments just by observing the average RGB value(R+G+B/3) of that image. It is found that almost in all cases the error values are lying between −10% to +10% of the actual value calculated from the flow rates of air and fuel. Diffusion flame is examined by passing fuel through the central channel of co-flow burner and air through the outer cylindrical channel. Air is used to stabilize the flame and for giving it a steady shape. Experiment is done keeping air flow constant while the fuel flow is varied and the image is captured. For diffusion flame, as the change in colour of flame is not much differentiable with the change in fuel, analysis is done to find the relation between fuel flow rate and flame area by counting the number of pixels. Finally, a direct relation of fuel amount and the image area is obtained.