Patent
Characteristic learning-based single image defogging method
Meihua Wang,Liang Yun,Jiaming Mai +2 more
- 29 Jun 2016
10
TL;DR: In this article, a multi-scale extraction of texture structure characteristics of a foggy image is performed via a sparse autoencoder, and at the same time, various color characteristics related with fog are extracted.
read more
Abstract: The invention discloses a characteristic learning-based single image defogging method. Firstly, multi-scale extraction of texture structure characteristics of a foggy image is performed via a sparse autoencoder, and at the same time, various color characteristics related with fog are extracted. Then, a multi-layer neural network is adopted for sample training, the mapping relationship between the texture structure characteristics and the color characteristics and the scene depth in the foggy condition is learned, and a scene depth chart of the foggy image is estimated. On this basis, a transmittance chart is approximately estimated by using the scene depth chart. The transmittance chart effectively reflects the fog concentration of each local area in the foggy image. Finally, with the combination of an atmospheric scattering model, restoration is further carried out according to the transmittance chart and a fogless image is obtained. The invention allows restoration of foggy images so as to obtain high-quality fogless images. In addition, compared with conventional defogging methods, the characteristic learning-based single image defogging method of the invention achieves better universal scene adaptability.
read more
Chat with Paper
AI Agents for this Paper
Find similar papers on Google Scholar, PubMed and Arxiv
Write a critical review of this paper
Analyze citations of this paper to find unaddressed research gaps
Citations
Patent
Single-image defogging method based on ResNet neural network
Wang Weixing,Huang Dewei,Jiang Bing,Chen Kexin,Lu Jianqiang +4 more
- 29 Jun 2018
TL;DR: In this article, a single image defogging method based on a ResNet neural network is proposed, which uses the ResNet network, can better extract features of the input image, has a very good defogging clearness effect on the foggy image under the specific scene, and is excellent in an image visual restoration effect.
10
Patent
Ranking convolutional neural network construction method and image processing method and device thereof
Xiaowu Chen,Song Yafei,Li Jia,Zhao Qinping,Xiaogang Wang +4 more
- 23 Nov 2016
TL;DR: In this paper, a ranking convolutional neural network (RCNN) is used to acquire output characteristics corresponding to an input characteristic pattern through automatic learning, compared with a method for acquiring characteristic through hand computation in the prior art, when applied to the field of image processing, the image processing effect can be greatly improved.
8
Patent
Image defogging algorithm on basis of dark channel prior
Lin Zhixian,Lin Shanling,Guo Tailiang,Ye Yun,Yang Bin,Shan Shengqi,Qian Mingyong,Zeng Suyun +7 more
- 06 Mar 2018
TL;DR: In this article, an image defogging algorithm on the basis of dark channel prior is proposed. But the method is not suitable for the case of foggy degraded images, and the quality of the images can not be improved.
7
Patent
Convolutional neural network defogging algorithm based on regional division and heavy fog preprocessing
Pang Yanwei,Lian Xuhang +1 more
- 27 Oct 2017
TL;DR: In this article, a convolutional neural network defogging algorithm based on regional division and heavy fog preprocessing is proposed, which comprises the following steps that: dividing a foggy image into non-overlapped image blocks, calculating the dark channel value of each image block, distinguishing a heavy fog image block and a mist image block; independently estimating transmissivity.
5
Patent
Single-image defogging method based on deep convolutional neural network
Wang Weixing,Huang Dewei,Lu Jianqiang,Lu Kexin +3 more
- 24 Apr 2018
TL;DR: In this paper, a single image defogging method based on a deep convolutional neural network is proposed, which comprises the steps of: acquiring an input image, and processing the input image via a pre-trained a rough depth-of-field image extraction network to obtain a rough DOF image; constructing a depth of field image refining network; feeding the rough depthoffield image and the input images which are as common inputs into the depth-ofthefield image refining networks, and carrying out estimation on an atmospheric scattering coefficient, solving out a transmittance image,
4
References
A Fast Single Image Haze Removal Algorithm Using Color Attenuation Prior
TL;DR: A simple but powerful color attenuation prior for haze removal from a single input hazy image is proposed and outperforms state-of-the-art haze removal algorithms in terms of both efficiency and the dehazing effect.
Back propagation neural network dehazing
Jiaming Mai,Qingsong Zhu,Di Wu,Yaoqin Xie,Lei Wang +4 more
- 01 Dec 2014
TL;DR: Experimental results show that the proposed approach is able to produce a high-quality haze-free image with the single hazy image and achieve the real-time requirement.
26
Patent
Single image defogging method and device on basis of BP (Back Propagation) neural network
Zhu Qingsong,Mai Jiaming,Wang Lei,Xie Yaoqin +3 more
- 08 Apr 2015
TL;DR: In this article, a single image defogging method was proposed based on a back propagation neural network (BP neural network) for obtaining fogless image information and calculating to obtain a corresponding fog image; by using RGB (Red Green Blue) values of pixel points of the fog image as inputs and using scene depths of the pixel points from the image as outputs, constructing a BP neural network model and generating a mapping relation between the pixel values and the depth values of fog image in the trained BP neural networks.
12
Patent
Remote sensing image change detection method based on sparse automatic code machine
Gong Maoguo,Qiunan Zhao,Ma Jingjing,Liu Jia,Li Hao,Ma Wenping,Mou Shugen,Hailun Yang +7 more
- 30 Dec 2015
TL;DR: In this paper, a remote sensing image change detection method based on a sparse automatic code machine was proposed for solving the disadvantages of too many undetected and wrongly detected pixels of a change detection result and incapability of using implicit information of a difference graph during direct processing of the difference graph in the prior art.
3
Related Papers (5)
Xu Wang,Qiudan Zhang,Jianmin Jiang +2 more
- 22 Feb 2017
Li Wenju,Chen Qi,Lu Yunfan,Hu Wenkang,Zhang Meng +4 more
- 27 Oct 2017