Imaging air quality evaluation based on Noise Brightness Ratio & Gompertz Type Diffusion Process

A novel air quality evaluation method, which only uses the surveillance image data, is presented in this paper. Since the intensity change of the Image Quality (IQ) can reflect the photography condition change of a CCD camera, the image can be used to evaluate the trend change of the air quality. In this paper, first a new IQ metric, which is nominated as the Noise Brightness Ratio (NBR), is defined. Then we use the NBR to build a Gompertz Type Diffusion Process (GTDP) of stochastic growth to fit the trend of that series. Once the GTDP is built, and if the selected NBR series is also a growth type, it can be used to evaluate and classify the IQ change level of different atmospheres. Some experiment data, which is captured in the period of dusk, are utilized to verify the validity of our method.