Spatially and Intensity Adaptive Morphology

Jean-Charles Pinoli (1, 2, 3, 4) , Johan Debayle (1, 2, 3, 4)

1 CIS-ENSMSE - Centre Ingénierie et Santé
2 IFRESIS - Institut Fédératif de Recherche en Sciences et Ingénierie de la Santé
3 STBio-ENSMSE - Surfaces et Tissus Biologiques
4 LPMG-EMSE - Laboratoire des Procédés en Milieux Granulaires

In this paper, spatially and intensity adaptive morphology is introduced and studied in the context of the General Adaptive Neighborhood Image Processing (GANIP) approach. The combination of GAN (General Adaptive Neighborhood)-based filtering and semi-flat morphology is particularly efficient in the sense that the filtering is adaptive to the image spatial structures (structuring elements are spatially variant) and its activity is controlled according to the image intensities (level sets are processed at different scales). The resulting morphological filters show a high image processing performance while preserving the image regions and details without damaging its transitions. The effectiveness of these adaptive operators are practically highlighted on real application examples for image background removing, image restoration and image enhancement.

Mots clés: Adaptive morphology; connected operators; general adaptive neighborhood image processing; generalized linear image processing; image filtering; semi-flat morphology; stack filtering

Domaines: Génie des procédés; Traitement du signal et de l'image [eess.SP]; Traitement du signal et de l'image [eess.SP]

Jean-Charles Pinoli, Johan Debayle. Spatially and Intensity Adaptive Morphology. IEEE Journal of Selected Topics in Signal Processing, 2012, 6 (7), pp.820-829. ⟨10.1109/JSTSP.2012.2214762⟩. ⟨hal-00837826⟩