In order to improve the visibility of intelligent ship vision system to sea fog images

aiming at the phenomena such as severe distortion of sky area

oversaturation and darkening of the restoration image caused by existing defogging algorithms when dealing with sea fog scene

an improved priori algorithm for sea fog removal is proposed. The depth map is obtained through color attenuation

and the dark channel map is obtained by compensating the dark channel with the median filtering. According to the depth of field map and dark channel map

the threshold value is set to segment the sky area

and the estimated atmospheric light value is obtained by analyzing the depth of field map of the sky area. Then

dark channel and color-attenuated dielectric transmission image obtain synthetic medium transfer diagram by wavelet fusion. To optimize defogging effect

medium transfer diagram are set suprathreshold and L-threshold for correction. By solving atmospheric scattering physical model

realize sharp processing of image. Finally

the problem of dark tone is optimized by Gamma correction. Through comparison experiments

it is verified that this method can obtain better quality and visual effects of defogging images in sea fog scenes.