对数域中基于实例学习的光照估计 下载: 994次
Illumination Estimation Based on Exemplar Learning in Logarithm Domain
合肥工业大学计算机与信息学院, 安徽 合肥 230009
图 & 表
图 1. 算法流程图
Fig. 1. Flow chart of algorithm
下载图片 查看原文
图 2. 两个场景在三组不同光照下的图像和对数色度直方图。场景1在(a)白光光照,(b)蓝光光照,(c)绿光光照下的图像和对数色度直方图;场景2在(d)白光光照,(e)蓝光光照,(f)绿光光照下的图像和对数色度直方图
Fig. 2. Images and log-chrominance histograms of two scenes with three different illuminations. Scene 1 images and log-chrominance histograms under (a) white light, (b) blue light, (c) green light; scene 2 images and log-chrominance histograms under (d) white light, (e) blue light, (f) green light
下载图片 查看原文
图 3. 单光照图像的光照估计流程
Fig. 3. Illumination estimation process of single illuminant image
下载图片 查看原文
图 4. 多光照图像的光照估计。(a)原始图像;(b)光照真实值;(c)单光照假设的估计结果;(d)双光照假设的估计结果;(e)多光照假设的估计结果
Fig. 4. Illumination estimation of multi-illuminant images. (a) Original images; (b) ground-truth values; (c) single illuminant estimation results; (d) double illuminant estimation results; (e) multi-illuminant estimation results
下载图片 查看原文
图 5. 不同光照估计方法在SFU Grey-ball数据集上的颜色校准结果。(a)原始图像;(b) Grey-World方法;(c) White-Patch方法;(d) Shades-of-Grey方法;(e) Grey-Edge方法;(f) Gamut Mapping方法;(g) Exemplar-Based方法;(h)本文方法
Fig. 5. Color correction results using different illumination estimation algorithms on SFU Grey-ball dataset. (a) Original images (b) Grey-World; (c) White-Patch; (d) Shades-of-Grey; (e) Grey-Edge; (f) Gamut Mapping; (g) Exemplar-Based; (h) proposed method
下载图片 查看原文
表 1图2中对数色度直方图χ2距离
Table1. χ2 distance of log-chrominance histograms in Fig. 2
χ2 distance | Fig. 2(a) | Fig. 2(b) | Fig. 2(c) | Fig. 2(d) | Fig. 2(e) | Fig. 2(f) |
---|
Fig. 2(a) | - | -12.5997 | -18.9088 | -3.6787 | -10.2847 | -17.0799 | Fig. 2(b) | -12.5997 | - | -21.6092 | -12.3675 | -3.6276 | -19.5451 | Fig. 2(c) | -18.9088 | -21.6092 | - | -18.0153 | -19.1128 | -3.6758 | Fig. 2(d) | -3.6787 | -12.3675 | -18.0153 | - | -9.4169 | -15.6009 | Fig. 2(e) | -10.2847 | -3.6276 | -19.1128 | -9.4169 | - | -16.7831 | Fig. 2(f) | -17.0799 | -19.5451 | -3.6758 | -15.6009 | -16.7831 | - |
|
查看原文
表 2不同光照估计方法在原始ColorChecker数据集上的角度误差对比
Table2. Angular errors for original ColorChecker dataset for different illumination estimation algorithms(°)
Method | Mean | Median | Trimean | Max |
---|
Do nothing | 6.9 | 9.5 | 7.5 | 38.2 | Grey-World | 9.8 | 7.4 | 8.2 | 46.0 | White-Patch | 8.1 | 6.0 | 6.4 | | Shades-of-Grey | 7.0 | 5.3 | 5.6 | 36.6 | Grey-Edge | 7.0 | 5.2 | 5.5 | | Zeta-Image | 6.9 | 5.0 | - | - | Gamut Mapping | 6.9 | 4.9 | 5.2 | 37.1 | Bayesian | 6.7 | 4.7 | | 39.4 | Weighted Grey-Edge | 6.6 | 4.7 | 5.1 | 44.3 | Exemplar-Based | | | - | - | Proposed algorithm | 5.1 | 3.4 | 3.8 | 28.5 |
|
查看原文
表 3不同光照估计方法在重处理的ColorChecker数据集上的角度误差对比
Table3. Angular errors for re-processing of ColorChecker dataset for different illumination estimation algorithms(°)
Method | Mean | Median | Trimean | Max |
---|
Do nothing | 13.7 | 13.6 | 13.5 | 27.4 | Grey-World | 6.4 | 6.3 | 6.3 | 24.8 | White-Patch | 7.6 | 5.7 | 6.4 | 40.6 | Shades-of-Grey | 4.9 | 4.0 | 4.2 | 22.4 | Grey-Edge | 5.1 | 4.4 | 4.6 | 23.9 | Zeta-Image | 4.1 | 2.8 | - | - | Gamut Mapping | 4.2 | 2.3 | 2.9 | 23.2 | Bayesian | 4.8 | 3.5 | 3.9 | 24.5 | Multi-Cue | 3.3 | 2.2 | 2.6 | - | Deep-CC | | | | 14.8 | Exemplar-Based | 2.9 | 2.3 | 2.4 | 19.4 | Proposed algorithm | 2.5 | 1.8 | 2.0 | |
|
查看原文
表 4不同光照估计方法在SFU Grey-ball数据集上的角度误差对比
Table4. Angular errors for SFU Grey-ball dataset for different illumination estimation algorithms(°)
Method | Mean | Median | Trimean | Max |
---|
Do nothing | 8.3 | 6.7 | 7.3 | 36.8 | Grey-World | 7.9 | 7.0 | 7.1 | 48.1 | White-Patch | 6.8 | 5.3 | 5.8 | | Shades-of-Grey | 6.1 | 5.3 | 5.5 | 41.2 | Grey-Edge | 5.9 | 4.7 | 5.1 | 41.2 | Gamut Mapping | 7.1 | 5.8 | 6.1 | 41.9 | Multi-Cue | 8.8 | 5.6 | 6.8 | - | Exemplar-Based | | | | 45.6 | Proposed algorithm | 4.2 | 3.0 | 3.4 | 43.0 |
|
查看原文
表 5不同光照估计方法在多光照数据集上的角度误差中位数对比
Table5. Median angular errors for multiple light sources dataset for different illumination estimation algorithms(°)
Method | Number of Illuminants |
---|
One | Two | Multi |
---|
Grey-World | 8.9 | 6.4 | - | White-Patch | 7.8 | 6.7 | - | Grey-Edge (n=1) | 6.4 | 5.6 | - | Grey-Edge (n=2) | | 5.1 | - | Exemplar-Based | 5.1 | | | Proposed algorithm | 4.2 | 3.6 | 3.7 |
|
查看原文
表 6本文方法在无分割和有分割设置下的角度误差对比
Table6. Angular errors for the proposed method with and without segmentation(°)
Dataset | Without segmentation | With segmentation |
---|
Mean | Median | Trimean | Max | | Mean | Median | Trimean | Max |
---|
Original ColorChecker | 5.3 | 3.7 | 4.0 | 32.5 | 5.1 | 3.4 | 3.8 | 28.5 | Re-processing of ColorChecker | 2.6 | 2.0 | 2.1 | 18.6 | 2.5 | 1.8 | 2.0 | 17.9 | SFU Grey-ball | 4.4 | 3.2 | 3.5 | 45.9 | 4.2 | 3.0 | 3.4 | 43.0 |
|
查看原文
表 7本文方法在无分割和有分割设置下的平均消耗时间对比
Table7. Average consuming time for the proposed method with and without segmentations
Dataset | Without segmentation | With segmentation |
---|
Estimation | | Segmentation+Estimation |
---|
Original ColorChecker | 0.9 | 6.3+89.3 | Re-processing of ColorChecker | 1.2 | 41.1+183.8 | SFU Grey-ball | 23.2 | 1.0+164.3 |
|
查看原文
崔帅, 张骏, 高隽. 对数域中基于实例学习的光照估计[J]. 光学学报, 2018, 38(2): 0233001. Shuai Cui, Jun Zhang, Jun Gao. Illumination Estimation Based on Exemplar Learning in Logarithm Domain[J]. Acta Optica Sinica, 2018, 38(2): 0233001.