advertisement
The paper describes restitution of geometrical distortions and improvement of signal-to-noise ratio of auto-fluorescence retinal images, finally aimed at segmentation and area estimation of the lipofuscin spots as one of the features to be included in glaucoma diagnosis. The main problems - geometrical and illumination incompatibility of frames in the image sequence and a non-negligible 'shear' distortion in the individual frames - have been solved by the presented registration procedure. The concept and some details of the MI-based regularized registration, together with evaluation of test results form the core of the contribution.
Dr. L. Kubecka, Department of Biomedical Engineering, Brno University of Technology, Brno, 61200, Czech Republic. kubecka@feec.vutbr.cz
6.9.5 Other (Part of: 6 Clinical examination methods > 6.9 Computerized image analysis)