VFX Project 1 - HDR Imaging

0. Team Members

• R07944007 林良翰
• R07922100 楊力權

1. Program Usage

1.1. Quick Usage

• Input directory

[input_directory]/
├──[image1]
├──...
└──shutter_times.txt

• Example of shutter_times.txt

32
16
8
4
2
1
1/2
1/4
1/8
1/16
1/32
1/64
1/128
1/256
1/512
1/1024

• Run

python3 hdr.py [input_directory] --savedir [output_directorp]    

1.2. Advanced Usage

• Please check the program document.

python3 hdr.py --help

2. Code Works

2.1. Image Alignment

• MTB Algorithm
  Instead of using median threshold bitmap, we used gradient magnitude to compare the differences.

2.2. HDR Reconstruction

• Debevec's Method
• Robertson's Method

2.3. Tone Mapping

• Photographic

  • Global Operator
  • Local Operator

• Bilateral We implemented opencv bilateral filtering to complete this algorithm.

3. Comparison

• Input Images - data/memorial

3.1. Debevec's Method

• Response Curve

  

• Radiance

  

• Tonemapping

Photographic Global Operator	Photographic Local Operator	Bilateral

3.2. Robertson's Method

• Response Curve

  

• Radiance

  

• Tonemapping

Photographic Global Operator	Photographic Local Operator	Bilateral

4. Our HDR Photo

• Input Images - data/ntu-library

4.1. Parameter

• Image Alignment

  • depth = $5$

• Debevec's Method

  • $\lambda = 10$

• Photographic Local Operator

  • $a=0.25$
  • $\delta=10^{-6}$

4.2. Result

• Response Curve - data/ntu-library/res/debevec/response_curve.png

  

• Radiance - data/ntu-library/res/debevec/radiance.hdr

  

• Tonemapping - data/ntu-library/res/debevec/tonemap_local.png