A method of image defogging realized on fpga

Abstract

The invention discloses an image defogging method realized on FPGA, comprising the following steps: inputting a color foggy image to the FPGA, obtaining pixel data of three channels of R, G, and B of the color foggy image and storing them in three channels respectively In a read-only memory with the same size; the pixel data of the three channels are respectively subjected to minimum value filtering, and the minimum value is taken to obtain the dark channel; the data of the dark channel and the three channels of R, G, and B are respectively thresholded The comparison method calculates the global atmospheric light of the three channels in parallel; calculates the transmittance according to the global atmospheric light, the simplified dark channel and the original image; restores the fog-free image according to the transmittance, atmospheric light and the original image. An image defogging method implemented on FPGA of the present invention solves the problem in the prior art that the processing speed of foggy images cannot meet the real-time processing requirements due to the slow processing speed.

Description

technical field [0001] The invention belongs to the technical field of image processing, and relates to an image defogging method realized on FPGA. Background technique [0002] In the technical fields of video surveillance, camera, navigation control, target tracking, etc., the processing of foggy images is involved. In severe weather such as fog and haze, the reflected light of objects is scattered and absorbed in the atmosphere, resulting in unclear images, loss of information, poor visibility, and color fading. Due to the particularity of the haze problem, at present, most of the images are dehazed by software through algorithms. Some algorithms have obtained good defogging effects, such as the dehazing algorithm that performs multi-scale Retinex processing on the brightness component, the bilateral filtering dehazing algorithm, and the dehazing algorithm based on the known depth of field model. These algorithms all require some specific The prior information is not co...

AI Technical Summary

Problems solved by technology

[0004] The purpose of the present invention is to provide an image defogging method implemented on FPGA, which solves the problem that the slow processing speed of foggy images in the prior art cannot meet the real-time processing requirements

Images