Design for Removing Haze Utilizing the Light Scattering Model and Dark Channel Prior Approach

Abstract

The atmospheric scattering effect will cause the picture taken to become blurry and partially gray and white in the frog and hazy climate. Therefore, it is important to understand the defogging algorithm in these conditions. For object protection and detection, the haze reduction process—which is carried out before data processing—is essential. It recovers a clearer picture from a hazy source picture. The computational efficiency of the numerous haze removal techniques that have been presented has to be improved. In order to produce an improved output image by using a hazy source image, a straightforward and effective haze removal technique that is applicable to VLSI hardware design was developed in this study. It is dependent on the dark channel previous method as well as the natural light scatter model. This work extracts daylight from a whole image. This method starts with the atmospheric diffusion model and generates an estimated emission map using a dark channel before combining it with grayscale to create a revised transmission map. Here, an approach for estimating the local atmospheric light is used in the design to improve the output result, as opposed to depending on a single global atmospheric light enabling the reconstruction of the blurred scene. In the final, the design performs better in both qualitative and quantitative evaluations without distorting or overloading the colors. A six-stage VLSI design has been suggested for the method in order to modify this process for a real-time application.