Image Enhancement

Description

This task provides a set of common image enhancing techniques to be applied to underwater images. It implements the color correction method in [Bianco2015], with the optional Contrast Limited Adaptive Histogram Equalization (CLAHE) step.

Configuration

The task can perform two image enhancement techniques: a color correction and a contrast stretching. The user can select using both or just one of them.

The color correction method does not have any parametrization. The user is referred to the original article [Bianco2015] to discern if the method is required for their data.

Regarding the contrast stretching, the CLAHE method is used. In this case, some parameters can be configured:

  • Tile grid size in X/Y: in CLAHE, the image is divided into small blocks and local histogram equalization is performed on them. These two parameters define the size of this partition.

  • Clip limit: in each of the tiles, CLAHE limits the amplification by clipping the histogram at this value.

Results

The enhanced images can be retrieved from the tasks list by clicking in the Download color corrected images button.

Additionally, there are other action buttons allowing to View the color corrected images, Compare the processed images with the input ones side by side or just View input image data.

Demo

Both datasets used in the mosaicing or the 3D reconstruction tasks can be used as sample input for this task.

Just upload the data to the task and wait for it to finish. Default options should be fine, but you can tune them to your convenience. Upon completion, click on the button labelled Compare the processed images with the input ones to see a quick online comparison between the results and the original data. You can also download the results by clicking on the Download color corrected images.

References

Bianco2015(1,2)

Bianco, G., Muzzupappa, M., Bruno, F., Garcia, R., and Neumann, L. A New Color Correction Method for Underwater Imaging Int. Arch. Photogramm. Remote Sens. Spatial Inf. Sci., XL-5/W5, 25–32. 2015. https://doi.org/10.5194/isprsarchives-XL-5-W5-25-2015