简介
本仓库提供了我的博士论文《基于原始响应值预测模型的数码相机图像信号处理方法与技术研究》相关代码及实验数据
数据
所有数据存储于 OneDrive，合计约129GB。若无特别说明则均含有 Nikon D3x 与 SONY ILCE7 两台相机的 raw 数据及实验结果
- 噪声标定（41.3GB）
- 原始响应值预测（78.6GB）
- 空间非均匀性标定（139MB，仅含 SONY D3x）
- 自动白平衡（8.2GB，仅含 SONY D3x）
- 颜色校正（1.2GB）
- 高光谱图像（352MB，收集自公开数据集）
测试环境
- Windows 10
- MATLAB R2018a/R2018b
MATLAB 所需工具箱
- Computer Vision System Toolbox
- Curve Fitting Toolbox
- Image Processing Toolbox
- Optimization Toolbox
- Parallel Computing Toolbox
- Statistics and Machine Learning Toolbox
更新日志
- 2020.05.18 - 上传数据
- 2019.12.18 - 定稿
- 2019.12.17 - 根据答辩组意见进行修改
- 2019.12.04 - 根据盲审意见进行修改
- 2019.07.04 - 完成论文初稿
目录结构
.
+---chapter1
+---chapter2
| +---colorimetry 色度学基础
| | \---figures
| \---radiometry_photometry 辐射度学、光度学基础
| \---figures
+---chapter3
| +---camera_noise
| | +---calibration 图像传感器噪声标定
| | | \---figures
| | \---correction 噪声修正
| | \---figures
| \---imaging_simulation_model
| +---parameters_estimation 原始响应值构成模型参数估计
| | +---comparisons
| | | +---PCA
| | | \---RBFN
| | \---figures
| \---preliminaries 成像系统非线性特性预实验
| \---figures
+---chapter4
| +---color_correction 颜色校正
| | \---figures
| +---metamer_mismatching 同色异谱与颜色失真
| | \---figures
| +---nonuniformity_correction 空间非均匀性校正
| | \---figures
| +---post_processing
| | \---chromatic_adaptation_transform 色适应变换后处理
| | \---figures
| \---white_balance_correction 自动白平衡（颜色恒常性）
| +---comparisons
| | +---figures
| | +---grayedge
| | \---weng
| +---gamut_mapping 2D色域映射算法
| | +---figures
| | \---utils
| +---neutral_pixels_statistics 中性色像素直方图统计算法
| | \---figures
| \---orthogonal_chromatic_plane 正交色度平面
+---chapter5
| +---luminance_estimation 场景亮度估计
| +---neutral_region 预设中性色区域选取
| | \---figures
| +---orthogonal_chromatic_plane 正交色度平面参数优化
| | \---figures
| \---standard_gamut 设备相关标准色域计算
| \---figures
+---chapter6
+---config 项目配置
+---data
\---utils
+---cie_diagram 色品图绘制
+---export_fig 矢量图导出工具
+---legendflex 图注工具
\---regtool 正则化工具
Intro
This repository contains codes for my doctoral thesis: Study on methodology and technology of digital camera image signal processing based on raw responses prediction model (zh-cn).
Data
All data is stored in OneDrive, with total size of 129GB.
- Noise calibration (41.3GB, Nikon D3x and SONY ILCE7)
- Response prediction (78.6GB, Nikon D3x and SONY ILCE7)
- Non-uniformity correction (139MB, only Nikon D3x)
- Auto white-balance (8.2GB, only Nikon D3x)
- Color correction (1.2GB, Nikon D3x and SONY ILCE7)
- Hyperspectral images (352MB, collected from public datasets)
Test Environment
- Windows 10
- MATLAB R2018a/R2018b
Required MATLAB Toolboxes
- Computer Vision System Toolbox
- Curve Fitting Toolbox
- Image Processing Toolbox
- Optimization Toolbox
- Parallel Computing Toolbox
- Statistics and Machine Learning Toolbox
ChangeLog
- 2020.05.18 - upload data
- 2019.12.18 - thesis completed
- 2019.12.17 - revised as per comments by examining committee
- 2019.12.04 - revised as per comments by anonymous reviewers
- 2019.07.04 - initial draft completed
License
Copyright 2019 Qiu Jueqin
Licensed under GPL-3.0.
React
Typescript
Django
Laravel
Facebook
Microsoft
Google
Alibaba
D3
Tencent