基于主动光照的深海图像增强算法 
下载： 609次
Deep Sea Image Enhancement Method Based on the Active Illumination
图 & 表
图 1. Underwater image formation model
Fig. 1. Underwater image formation model
下载图片 查看原文
图 2. Schematic of underwater light attenuation
Fig. 2. Schematic of underwater light attenuation
下载图片 查看原文
图 3. The flowchart of the proposed method
Fig. 3. The flowchart of the proposed method
下载图片 查看原文
图 4. Visual examples of recovered images with different background light estimations
Fig. 4. Visual examples of recovered images with different background light estimations
下载图片 查看原文
图 5. Segmentation of water and the object
Fig. 5. Segmentation of water and the object
下载图片 查看原文
图 6. The proposed background light estimation
Fig. 6. The proposed background light estimation
下载图片 查看原文
图 7. The flowchart of the color correction
Fig. 7. The flowchart of the color correction
下载图片 查看原文
图 8. Comparison of dehazed image and the final enhancement(dehazing and color correction)result
Fig. 8. Comparison of dehazed image and the final enhancement(dehazing and color correction)result
下载图片 查看原文
图 9. Comparison of color space
Fig. 9. Comparison of color space
下载图片 查看原文
图 10. Comparison of histogram
Fig. 10. Comparison of histogram
下载图片 查看原文
图 11. Subjective comparison of different enhancement methods on deep underwater scene images
Fig. 11. Subjective comparison of different enhancement methods on deep underwater scene images
下载图片 查看原文
表 1水下图像UCIQE和熵的客观比较
Table1. Objective comparison results of UCIQE and entropy
| UCIQE | Original image | DCP | DehazeNet | IBLA | SIDMP | COFFID | Proposed method | | Image1 | 0.326 1 | 0.339 2 | 0.403 5 | 0.418 1 | 0.419 2 | 0.402 3 | 0.460 2 | | Image2 | 0.275 4 | 0.378 6 | 0.311 2 | 0.357 0 | 0.344 8 | 0.367 3 | 0.539 1 | | Image3 | 0.306 5 | 0.352 6 | 0.309 6 | 0.333 3 | 0.353 7 | 0.403 6 | 0.503 5 | | Image4 | 0.196 7 | 0.187 3 | 0.224 3 | 0.274 0 | 0.224 7 | 0.253 3 | 0.431 0 | | Entropy | Original image | DCP | DehazeNet | IBLA | SIDMP | COFFID | Proposed method | | Image1 | 7.032 5 | 7.047 6 | 7.113 2 | 7.526 6 | 7.028 4 | 7.102 6 | 7.701 2 | | Image2 | 6.985 3 | 7.101 6 | 6.998 3 | 7.362 1 | 7.393 6 | 7.035 1 | 7.795 2 | | Image3 | 7.179 5 | 7.283 0 | 7.256 5 | 7.228 5 | 7.229 8 | 7.367 9 | 7.897 8 | | Image4 | 6.190 4 | 6.111 1 | 6.443 0 | 6.767 2 | 6.432 7 | 6.273 5 | 7.994 4 |
|
查看原文
邓翔宇, 王惠刚, 张永庆. 基于主动光照的深海图像增强算法[J]. 光子学报, 2020, 49(3): 0310001. Xiang-yu DENG, Hui-gang WANG, Yong-qing ZHANG. Deep Sea Image Enhancement Method Based on the Active Illumination[J]. ACTA PHOTONICA SINICA, 2020, 49(3): 0310001.
PDF全文
