AlignExposures |
The base class for algorithms that align images of the same scene with different exposures
|
AlignMTB |
This algorithm converts images to median threshold bitmaps (1 for pixels brighter than median
luminance and 0 otherwise) and than aligns the resulting bitmaps using bit operations.
|
CalibrateCRF |
The base class for camera response calibration algorithms.
|
CalibrateDebevec |
Inverse camera response function is extracted for each brightness value by minimizing an objective
function as linear system.
|
CalibrateRobertson |
Inverse camera response function is extracted for each brightness value by minimizing an objective
function as linear system.
|
MergeDebevec |
The resulting HDR image is calculated as weighted average of the exposures considering exposure
values and camera response.
|
MergeExposures |
The base class algorithms that can merge exposure sequence to a single image.
|
MergeMertens |
Pixels are weighted using contrast, saturation and well-exposedness measures, than images are
combined using laplacian pyramids.
|
MergeRobertson |
The resulting HDR image is calculated as weighted average of the exposures considering exposure
values and camera response.
|
Tonemap |
Base class for tonemapping algorithms - tools that are used to map HDR image to 8-bit range.
|
TonemapDrago |
Adaptive logarithmic mapping is a fast global tonemapping algorithm that scales the image in
logarithmic domain.
|
TonemapMantiuk |
This algorithm transforms image to contrast using gradients on all levels of gaussian pyramid,
transforms contrast values to HVS response and scales the response.
|
TonemapReinhard |
This is a global tonemapping operator that models human visual system.
|