The evaluation of air quality using image quality
利用图像质量评估空气质量的新方法

The intensity change of the image quality (IQ) can reflect the photography condition change of a CCD camera. A novel air quality evaluation method, which only uses the surveillance image, is presented in this paper. First, different IQ metrics, i.e. the noise brightness ratio, the contrast level and the blur level, are employed to evaluate the IQ. Then we use them to build different time series data. After that we utilize a Gompertz type diffusion process (GTDP) of stochastic growth and the detrend fluctuation analysis (DFA) method to analyze and fit the trend of these series above. Finally, we can use the parameters of the GTDP model and the DFA model to evaluate and classify the IQ change level of different atmospheres. Some experiment data, which are captured in a solar day and a misty day, are utilized to verify the validity of our method.