Research And Application Of Image Processing In Harsh Environment

Haze weather not only harms human's health, but also will seriously affect the work of the computer vision system because the fog will make the suspension of particles in the air making the light scattering effect which causes unclear image acquisition in a foggy day, the problem of low contrast, etc. So the image contrast narrated technology research is a hot issue in the f-ield of computer vision. Based on the analysis of the formation of haze weather , the introduction of atmospheric scattering model and the study of the fog image degradation model, the main research content is as follows:Firstly, it focuses on the image defogging algorithm, based on the dark channel priority and the image defogging algorithm based on the optimization of contrast enhancement. The image defogging algorithm of the dark channel priority is a kind of methods to gain prior information that is firstly according to the statistical knowledge,then recuperates image with the help of atmospheric scattering model. The image defogging algorithm based on the optimization of contrast enhancement is a kind of image enhancement methods that balances image contrast and information loss.Secondly, aiming at the image distortion issue in the region of the sky after using the image defogging algorithm based on the dark channel priority, an adaptive method to improve the tolerance mechanism is adopted. by introducing the tolerance of K, the transmission rate of the image divided into bright area and normal area is optimized,so as to make an improvement on the sky color distortion problem while recuperating image.Then based on the Matlab platform, the images dividing into the mist and fog figures are conducting simulation experiments using the methods mentioned in this paper. Through experiments, it shows that the image defogging algorithm based on the dark channel priority is suitable for the mist image while the image defogging algorithm based on the optimization of contrast enhancement is suitable for fog image.This paper also chose two pictures taken under different weathers as experimental objects, in order to demonstrate its point of view and made a comparison between the two algorithms and classic algorithm in the field of image restoration and image enhancement using subjective evaluation and objective evaluation, with objective evaluation mainly adopting the image evaluation method based on the visible edge contrast enhancement.It is proved byexperimnet that this algorithm has better stability.Finally, applying the image defogging algorithm based on the optimization of contrast enhancement which is studied in this paper to the .image obtained by unmanned aerial vehicle in a fire rescue, whose feasibility of the application is demonstrated after the experiment.