A feature-based dense local registration of pairs of retinal images.

Mathieu Fernandes (1, 2, 3, 4) , Yann Gavet (1, 2, 3, 4) , Jean-Charles Pinoli (1, 2, 3, 4)

1 CIS-ENSMSE - Centre Ingénierie et Santé
2 DIM-ENSMSE - Département Imagerie et Morphologie
3 LPMG-EMSE - Laboratoire des Procédés en Milieux Granulaires
4 IFRESIS - Institut Fédératif de Recherche en Sciences et Ingénierie de la Santé

A method for spatial registering pairs of digital images of the retina is presented, using intrinsic feature points (landmarks) and dense local transformation. First, landmarks, i.e. blood vessel bifurcations, are extracted from both retinal images using filtering followed by thinning and branch point analysis. Correspondances are found by topological and structural comparisons between both retinal networks. From this set of matching points, a displacement field is computed and, finally, one of the two images is transformed. Due to complex retinal registration problem, the presented transformation is dense, local and adaptive. Experimental results established the effectiveness and the interest of the dense registration method.

Mots clés: Feature-based registration; Retinal images; Opthalmology; Local transformation; Dense transformation.; Dense transformation

Domaines: Génie des procédés

Mathieu Fernandes, Yann Gavet, Jean-Charles Pinoli. A feature-based dense local registration of pairs of retinal images.. VISAPP 2009: 4th International Conference on Computer VISion Theory and APplications, Feb 2009, Lisbonne, Portugal. pp.265-265. ⟨hal-00509312⟩