Research On Image Defogging Algorithm Based On Atmospheric Scattering Model

With the development of electronic circuit technology and optical imaging technology,the image resolution and clarity are getting better and better.However,in harsh weather and other adverse weather conditions,the image will be seriously degraded,and bring great negative impact on the aerospace,industrial production,transportation and biomedical.The main cause of image degradation is that the target light interacts with the suspended particles in the air during propagation,such that image detail is lost,hue shift,desaturate thus can not meet the requirements of human vision.In order to improve the stability and applicability of outdoor imaging system,this paper presents an algorithm based on image defogging atmospheric scattering model.Experiments show that the new algorithm has strong applicability and high stability.The main contents are as follows:The traditional polarization defogging algorithm considers the degree of polarization as a global invariant,caused that the overall image dark,low gradation,low tones and loss of detail information of the image.To overcome these disadvantages,we propose a polarization image defogging algorithm based on mean filtering.First we should get two polarization images on both the best and the worst status using polarization imaging detection system,then estimate the Atmospheric light intensity by the quad-tree subdivision method.So we can fulfill the Automatic optimization of the light intensity of atmosphere and other related parameters and estimate the Atmospheric Scattered Light by means of the deformation form of the mean filtering.At last,we restore the image based on the model of defogging so that we can realize the defogging of the image.The experiment results found this algorithm suitable for many scenes,with higher contrast,more clear image details and richer tones.So it has a certain visual advantage.We propose a fast fog image restoration algorithm based on the total variation when we considering the deficiency of the fast defogging algorithm based on the median filtering.We estimate the atmospheric dissipation function with the total variation first.It avoids the phenomenon of Edge fog and partial darkness.Then we introduce the adaptive protection factor under the Histogram correction mechanism to correct the atmospheric scattering function of the bright area.At the same time,the He algorithm is used to obtain the atmospheric light value quickly and accurately through the dark channel.Thus reduce the time complexity while avoiding the impact of the highlighted object on the A value estimate in the fogging image.At last,we adjust the brightness of the reconstructed image to attain a effect picture with richer color and more clear details.Moreover,the time complexity of the algorithm is a linear function of the number of image pixels so that it increases the calculation speed significantly.