Research On Image Defogging Technology And Application

With the rapid development of industrialization and urbanization,haze weather has become more and more frequent,and the turbid substances in the air gradually increase,resulting in a decrease in the definition of captured images which are hard to distinguish.However,with the development and progress of science and technology,the video systems based on image vision have been widely used,such as driving recorders,real-time navigation systems and outdoor monitoring equipment,which have played an important role in improving the quality of living standards and the efficiency of production and ensuring the safe and stable operation of society.Therefore,the study of high-quality,fast and universal image defogging technology has great practical significance and theoretical value.This article combines the current single-image defogging technology for research,and the main work is as follows:Firstly,an image defogging method based on sky segmentation is proposed,which combines sky segmentation and image enhancement methods to effectively image defog.The image is performed by mixed processing of secondary sharpening and smoothing to preserve the detailed information and reduce the distortion.In addition,the Ostu threshold segmentation algorithm is used to divide the foggy image into the sky area and the non-sky area,which are processed by different defogging methods.Finally,the defogging images of the sky area and the non-sky area are fused to output a complete defogging image.Secondly,combining the quadtree algorithm and the adaptive linear method,an improved dark channel defogging algorithm is proposed.On the one hand,an improved quadtree algorithm is used to determine the atmospheric light value,which effectively improves the accuracy of atmospheric light value;on the other hand,an improved adaptive linear method is adopted to estimate the projection rate,which effectively enhances the adaptability to the sky area.The defogging image is more suitable for human vision,and the processing efficiency of the algorithm is also effectively improved.Finally,a camera system based on image defogging is designed to solve out the problem of low definition of images taken under outdoor haze weather.The system consists of a fog concentration detection module,an image defogging module and an object recognition module,which can effectively process and recognize the image of input,which can greatly improve the efficiency of work.