Research On Haze Filtering Of Aerial Images

The aerial images with fog have disadvantages of low contrast,fuzzy scenes and missing image information,so it not only reduces greatly the value of the images and the accuracy of the map in the industry.With the development of computer vision technology,we have the ability to obtain the recovery and reduction of low information with fog to rich information without fog by a certain image processing method.Therefore,how to improve and enhance the degradation of image quality caused by the haze has become a new hot spot in the field of aerial photogrammetry.In this paper,we do the following works to investigate hardware and software system in order to remove the haze of aerial images.1)An improved algorithm based on dark channel priorIn terms of software algorithms,the paper Mainly introduces Image enhancement defogging method of histogram equalization,Retinex theory and the dark channel prior algorithm based on image restoration.Then it analyses and compares the characteristics and problems of these methods from the theoretical basis,process efficiency,color fidelity and other aspects.In the paper,the dark channel prior algorithm based on image degradation model is chosen for this study.In addition,according to the characteristics of the aerial images and the shortages of dark channel prior algorithm,introducing the automatic white balance algorithm,using alternating sequence to replace soft matting in the original algorithm to reduce the processing time,the optimization algorithm is proposed in this paper.The experimental results indicate that the optimization method is better to the fog,image color fidelity and high efficiency.2)The defogging method based on infrared image and visible image fusionIn terms of hardware,it researches and develops an integrated camera that can get the visible and infrared images simultaneously through the transformation process of Hasselblad single based on infrared light filtering glass and reinforcement technology.The dual image fusion method to defog is proposed based on stronger penetrating of infrared than visible light.The experimental results show that this method has good feasibility and effectiveness.3)The program realization of the two methodsAccording to the Matlab software development platform,we have designed simple program system of the optimization algorithm and fusion defog method using Graphical User Interfaces(GUI)tool.