Design of an FPGA-based Image Defogging Algorithm Using Dark Channel Prior

In recent years, with the rapid advancement of digital image processing technology, image dehazing has become a focal point in both scientific research and practical applications. This paper presents a study on a dehazing algorithm based on the dark channel prior and its optimization for FPGA implementation. To improve image refinement and reduce algorithm complexity, the filter window size for calculating the dark channel is simplified to 3×3. When estimating the global atmospheric light intensity (A), the brightest pixel in the dark channel image is selected as the A value. Based on this, the transmission map is computed, and a guided filter is introduced to smooth the image. Finally, FPGA is utilized to accelerate the algorithm, and a low-pass filter is employed instead of the guided filter to further reduce complexity and enhance processing efficiency, meeting the real-time dehazing requirements for high-resolution images.