| Summary: | The haze weather has seriously influenced the image quality of outdoor video systems. With the extensive and in-depth application of outdoor video systems, there is an urgent need for miniaturized and embedded real-time image defog systems. This paper proposes a real-time image defog method based on ZYNQ, in which the luminance component V is processed by the Retinex algorithm in the HSV color space. With the cooperation of software calculation and hardware operation, the algorithm control and simple operations, such as image color space conversion and logarithm operations, are performed in the ARM. In order to improve the speed of the algorithm, the parallel convolution algorithm between the Gaussian kernel function and two-dimensional image is executed in FPGA to estimate the ambient light illumination. The experimental results have shown that the proposed algorithm has the advantages of high speed, miniaturized, embedded, high portability and low power consumption, fulfilling the practical requirements of outdoor video systems.
|
|---|
