haader
======

A tool to construct HDR-images from a series of photographs of the same scene
with different exposures. It reads JPEG-files and parses the EXIF data to
retrieve their exposure times. The sensor response function is estimated using a
method which is similar to the one described by Debevec and Malik in 1997.

([Paul E. Debevec and Jitendra Malik. "Recovering High Dynamic Range Radiance
Maps from Photographs". In Proceedings of SIGGRAPH 97, Computer Graphics
Proceedings, Annual Conference Series, pages 369–378,
1997.](http://www.pauldebevec.com/Research/HDR/debevec-siggraph97.pdf))


There are two interfaces available.

## Command Line Interface

To build the CLI, you will just need a C++ compiler as all dependencies are
included:
```sh
# apt-get install build-essential
```

Build:
```sh
$ cd cli
$ make
```

Run:
```sh
$ cd cli
$ ./haader <image1.jpg> <image2.jpg> <image3.jpg> ...
```

This will output several files if the reconstruction of the sensor response
function was successful:

* `expose_*.ppm`: Simulated exposures of the scene with different exposure
  times
* `log_image.ppm`: A logarithmic mapping of the HDR-image
* `rf.csv`: The sensor response function as a simple table with two columns. The
  first column contains the logarithmic luminance relative to the central
  response (128). The second row contains the response value (0-255).
* `irf.csv`: The inverse sensor response function as a simple table with two
  columns.

## GTK Interface

Needs GTK+3.0 and the corresponding development packages:
```sh
# apt-get install build-essential libgtk-3-0 libgtk-3-dev
```

Build and run:
```sh
$ cd gui_gtk
$ make run
```

Click `Select files` to open multiple images of the same scene and start the
reconstruction. Alternatively, it is possible to specify the images and exposure
times with a simple two-column CSV file:
```csv
</absolute/image/path1.png>,<exposure time in seconds>
</absolute/image/path2.png>,<exposure time in seconds>
...
```