Abstract: The image dehazing algorithm restores low-quality images captured in hazy environments to clear images. However, with the increase in image resolution and algorithm complexity, it is challenging to ensure the real-time performance of the algorithm in practical applications. To address this issue, this paper studies and improves a low-complexity ROP (rank-one prior) algorithm and enhances its execution efficiency through an FPGA（field programmable gate array） hardware platform. Firstly, leveraging the parallel processing advantages of FPGA, the algorithm incorporates spatial correlation and dark channel prior constraints to eliminate interference from close-range and high-brightness areas. By optimizing the scattering rate map estimation method, the improved algorithm resolves the artifacts present in images restored by the original ROP algorithm while reducing hardware resource consumption. Finally, the clear image is obtained by solving the estimated ambient light value and scattering rate map. Experimental results demonstrate that the improved algorithm enhances the visual quality of images in scattering scenes such as fog and underwater environments. The restored images exhibit more realistic colors and more details. Implementing this algorithm on the ZYNQ7020 development board (utilizing 21K LUT and 28.9% BRAM resources) processes 1080p images in 54 ms, meeting real-time processing requirements. This has broad applications in fields like autonomous driving and deep-sea exploration.