Motivation:

Multi-modality registration is a process of combining information from multiple images to generate a single image that contains a more accurate description of the scene than any of the individual source images. FSome structures visible in one modality may not be seen in the other imaging modalities. Multisensor data often presents complementary information about region surveyed, scene or object so image registration provides an effective method to evaluate different aspects of the sane scene. Aligning two different modality images of the same scene is common in image fusion systems where significant image misalignments are removed through transformation of the one image onto other. Image fusion may be regarded as a further step after multimodality registration in which data integration is implemented to obtain a composite image.

Many strategies have been proposed and implemented for the image registration based on either the geometrical features (point-like anatomic features or surfaces) or intensity similarity measures. Knowledge of a 1-to-1 relationship between the grey value images is used in intensity similarity measures.

Objectives:

The goal of this project is to register Face Images to make a mosaic image.

Technical Approach:

Transformation matrices that maps coordinates between face pairs is investigated using MI based registration. Nelder-Mead downhill simplex method is used for optimization to accelerate the algorithm. After aligning each face image to the their exact positions a very simple linear fusion algorithm is applied

Results:

Currently in process

Publications:

No publications currently available for this project.

This research is being conducted at the IRIS Lab by Muharrem Mercimek under the supervision of Dr. Mongi A. Abidi.