A single image defogging method based on a convolutional neural network

Abstract

The invention provides a single image defogging method based on a convolutional neural network. The method comprises the steps that firstly, a training set is constructed to serve as input of a deep convolutional neural network model, the network model comprises a shallow neural network model and a deep neural network model, and the shallow neural network model is used for extracting and fusing features of RGB color space of a foggy image and outputting a scene depth map of the foggy image; and the deep network model performs multi-scale mapping, pooling, convolution and other operations on the scene depth map on the basis of the shallow network model, and outputs a transmissivity map of the foggy image. And finally, recovering the fog-free image through the transmissivity, the atmosphericlight value and the atmospheric scattering model. According to the method, the characteristics of the RGB color space of the atomized image are extracted and fused to construct the shallow convolutional neural network model, and the shallow convolutional neural network model is connected with the multi-scale deep neural network model to establish the end-to-end neural network model, so that defogging clearness can be realized in various scenes, and particularly, color distortion of the image can be avoided in a dark environment.

Description

technical field [0001] The invention relates to a method for defogging a single image, in particular to a method for defogging a single image based on a convolutional neural network. Background technique [0002] Due to garbage incineration, construction dust, vehicle exhaust emissions and other reasons, many cities in China have cast the shadow of smog. Smog has become an environmental problem that has received continuous attention in recent years. Due to the decrease of contrast and color saturation of images taken in foggy weather, the pictures are not clear, which affects the use of pictures. For example, traffic surveillance video shooting in foggy days is blurred, resulting in deviations in the image recognition and processing process, which is not conducive to accurate recording of traffic information. Therefore, there are urgent theoretical and practical needs to improve the image quality in foggy weather and reduce the impact of foggy weather on outdoor imaging. ...

AI Technical Summary

Problems solved by technology

[0005] Although the dehazing algorithm based on image restoration is relatively effective, it is not universal because the simplified physical model is based on the condition that the atmosphere is single-scattering and the medium is uniform, such as uneven fog or sky areas, and in dark areas. Easy to cause color distortion in the environment

Images