A real-time image dehazing method considering dark channel and statistics features

In recent years, image dehazing algorithms are promoted, but they have not been used in real-time processing. This paper proposed a combined algorithm based on both dark channel prior and histogram optimization. First of all, the histogram optimization algorithm are used in image preprocessing, which can make the image contrast stretching, so the impact of the haze on the image can be weakened. If the obtained dehazing image can meet the requirements of the system, it will no longer be dealed with in following treatment, so we can save a lot of processing time. If it cannot meet the requirements, the dark channel prior can be used to estimate the haze intensity. According to the characteristics of the haze image, the correlation in frequency domain can be chosen. In this way, the software system can quickly deal with the images or videos to achieve real-time application requirements. Experiments show that proposed algorithm can not only meet the basic requirements for image dehazing, but also can improve the computational efficiency, so as to meet the application of real-time image processing.