LIDRIC is a "A Local Image Descriptor Robust to Illumination Changes" and is based on the paper of the same title presented at SCIA 2013. Experiments show the superiority of this descriptor compared to others, e.g., SIFT, DAISY, SURF, FREAK.
The current implementation is fully written in Matlab and is freely available for non-commercial research use. Please note that the code has not been fully runtime optimized, although version 1.02 brings a considerable speed-up compared to version 1.0. You can check "main.m" for how to use the code.
Please cite our paper when using the code for your research:
Zambanini S., Kampel M. "A Local Image Descriptor Robust to Illumination Changes", Proc. of Scandinavian Conference on Image Analysis - SCIA 2013, pp. 11-21, 2013.
LIDRIC can be used for feature-based image registration, see for instance http://de.mathworks.com/help/images/examples/find-image-rotation-and-scale-using-automated-feature-matching.html
You only have to change the extractFeatures-step to LIDRIC feature extraction.
would be nice if you provide simple example for image registration ;)
the paper and code are both interesting，but there is no example for application.and i think the feature dimension is too high.
Any application where you rely on local image features and strong illumination variations can occur between images, e.g. image registration, object recognition...
Thank you sharing this code. What are the different applications of this code?
Corrected a small bug in "FeatureMap.m"
Changed the description.
Download apps, toolboxes, and other File Exchange content using Add-On Explorer in MATLAB.