Chinese Journal of Liquid Crystals and Displays, Volume. 38, Issue 2, 225(2023)
Dehazing algorithm of monochromatic remote sensing image based on matrix restoration and dark channel theory
Figures & Tables(8)
Fig. 4. Restoration and dehazing results of 30% information loss haze image
Fig. 5. Restoration and dehazing results of images with different proportions of information lost
Table 1. Calculation results of objective evaluation index for image dehazing
View tableView in ArticleTable 1. Calculation results of objective evaluation index for image dehazing
农田 城市 建筑 海洋 河流 山脊 平均值 雾图 对比度 18.612 1 16.873 3 21.912 6 37.751 7 33.995 2 26.447 9 25.932 1 峰值信噪比 14.251 2 11.530 0 11.300 7 14.172 5 18.963 9 9.578 2 13.299 4 结构相似性 0.869 3 0.864 0 0.705 6 0.819 5 0.886 5 0.585 7 0.788 5 算法1 对比度 55.084 0 37.508 3 46.515 0 60.776 6 63.458 5 53.047 1 52.731 6 峰值信噪比 9.105 3 9.666 1 9.429 5 10.342 2 19.256 4 7.069 5 10.811 5 结构相似性 0.668 2 0.846 6 0.777 7 0.768 7 0.894 2 0.587 5 0.757 2 算法2 对比度 29.790 0 15.329 2 20.744 4 53.051 8 39.560 4 35.197 6 32.278 9 峰值信噪比 16.261 9 14.316 8 13.185 0 17.003 0 20.550 9 11.288 0 15.434 3 结构相似性 0.794 0 0.865 2 0.684 7 0.876 4 0.894 4 0.670 2 0.797 5 算法3 对比度 65.356 5 56.228 5 61.903 9 73.298 5 76.734 2 70.474 4 67.332 7 峰值信噪比 10.656 9 13.255 5 15.756 8 14.503 7 17.104 3 10.509 2 13.631 1 结构相似性 0.643 9 0.878 8 0.903 3 0.839 3 0.826 9 0.711 0 0.800 5 算法4 对比度 39.274 6 25.216 6 58.794 9 63.444 7 55.650 8 50.968 8 48.891 7 峰值信噪比 19.465 3 24.347 9 19.392 8 19.095 0 11.940 3 14.914 4 18.192 6 结构相似性 0.798 8 0.927 3 0.860 0 0.892 1 0.802 5 0.796 3 0.846 2 算法5 对比度 38.191 4 37.332 2 51.275 6 58.131 3 65.198 2 48.849 8 49.829 8 峰值信噪比 17.533 9 14.923 2 15.058 2 17.523 2 22.329 5 11.753 8 16.520 3 结构相似性 0.873 0 0.933 1 0.897 7 0.904 7 0.910 2 0.757 7 0.879 4 算法6 对比度 18.612 1 28.193 6 35.674 7 44.559 3 56.811 1 37.914 2 36.960 8 峰值信噪比 14.251 2 12.642 2 14.757 7 16.779 3 24.451 5 10.102 4 15.497 4 结构相似性 0.869 3 0.907 3 0.860 4 0.889 1 0.976 1 0.684 7 0.864 5 本文算法 对比度 29.479 5 31.752 5 41.431 7 48.692 3 58.025 2 34.704 7 40.681 0 峰值信噪比 22.981 0 20.477 3 19.814 0 26.819 3 21.589 0 17.789 1 21.578 3 结构相似性 0.950 5 0.972 8 0.941 5 0.971 2 0.946 9 0.842 9 0.937 6
Table 2. Calculation results of objective evaluation indexes for image restoration and dehazing
View tableView in ArticleTable 2. Calculation results of objective evaluation indexes for image restoration and dehazing
农田 城市 建筑 海洋 河流 山脊 平均值 信息丢失雾图 信息熵 5.246 9 5.109 1 5.369 0 5.940 8 5.392 1 5.595 8 5.442 3 峰值信噪比 12.048 1 8.946 4 9.022 1 10.819 2 7.742 2 9.417 7 9.666 0 结构相似性 0.054 1 0.013 7 0.058 4 0.115 8 0.038 2 0.045 6 0.054 3 算法1 信息熵 5.157 5 4.863 5 5.086 3 5.703 1 5.036 5 5.401 5 5.208 1 峰值信噪比 7.546 4 7.213 2 6.852 6 8.075 6 7.175 1 6.362 9 7.204 3 结构相似性 0.039 0 0.010 9 0.045 6 0.083 9 0.027 1 0.030 2 0.039 5 算法2 信息熵 5.397 0 5.448 9 5.812 9 6.041 3 5.808 4 5.796 2 5.717 5 峰值信噪比 12.467 4 9.300 7 9.394 1 11.249 3 7.784 8 9.912 0 10.018 0 结构相似性 0.062 0 0.015 6 0.072 7 0.136 9 0.045 9 0.055 4 0.064 7 算法3 信息熵 5.165 9 5.036 1 5.274 0 5.770 5 5.301 6 5.460 5 5.334 8 峰值信噪比 10.088 4 9.184 9 10.088 1 11.443 8 7.137 9 10.430 4 9.728 9 结构相似性 0.129 1 0.042 2 0.174 6 0.247 3 0.111 2 0.193 7 0.149 7 算法4 信息熵 6.604 3 6.736 3 6.969 6 7.259 3 7.288 3 7.096 9 6.992 5 峰值信噪比 12.422 4 13.682 7 10.979 2 11.787 8 10.473 5 10.123 3 11.578 2 结构相似性 0.063 3 0.032 5 0.073 2 0.131 6 0.066 2 0.058 9 0.070 9 算法5 信息熵 5.398 3 5.223 3 5.531 0 5.985 2 5.462 7 5.731 8 5.555 4 峰值信噪比 10.887 1 8.076 2 7.920 9 10.659 0 7.511 8 8.396 7 8.908 6 结构相似性 0.050 9 0.012 6 0.058 0 0.117 6 0.037 1 0.043 2 0.053 2 算法6 信息熵 5.3433 5.194 0 5.302 5 5.939 4 5.598 1 5.595 8 5.495 5 峰值信噪比 10.658 8 8.925 6 9.127 4 10.822 3 7.783 2 8.760 6 9.346 3 结构相似性 0.046 0 0.014 1 0.060 2 0.114 2 0.041 3 0.041 9 0.052 9 本文算法 信息熵 6.778 5 6.921 0 7.116 8 7.483 8 7.207 7 7.137 2 7.107 5 峰值信噪比 22.534 9 19.782 1 19.739 2 26.358 7 23.207 2 16.591 4 21.368 9 结构相似性 0.855 7 0.930 6 0.844 9 0.951 3 0.894 1 0.737 2 0.868 9
Table 3. Calculation results of objective evaluation indexes for restoration and dehazing of fog images with different proportions of information loss
View tableView in ArticleTable 3. Calculation results of objective evaluation indexes for restoration and dehazing of fog images with different proportions of information loss
图像信息比例 信息熵 峰值信噪比 结构相似性 50%图像信息 4.177 3 8.879 0 0.019 8 7.104 8 11.516 9 0.529 5 60%图像信息 4.783 2 8.602 1 0.021 1 7.105 9 11.880 5 0.591 3 70%图像信息 5.329 5 8.334 3 0.02 7 7.113 1 11.888 8 0.596 1 80%图像信息 5.807 1 8.079 3 0.035 4 7.134 5 11.678 6 0.596 4
Tools
Get Citation
Copy Citation Text
Ji-wei LIU, Xiao-dong WANG, Yun-hui LI. Dehazing algorithm of monochromatic remote sensing image based on matrix restoration and dark channel theory[J]. Chinese Journal of Liquid Crystals and Displays, 2023, 38(2): 225
Paper Information
Category: Research Articles
Received: Jun. 8, 2022
Accepted: --
Published Online: Feb. 20, 2023
The Author Email: Xiao-dong WANG (wangxd@ciomp.ac.cn)
© Copyright 2018-2021 | Chinese Laser Press.
All Rights Reserved 沪ICP备15018463号-20