Manhole cover detection using a geometrical filter on very high resolution aerial and satellite images

Détection automatique de regards du réseau d'assainissement par analyse d'images satellites

Olivier Bartoli (1) , Nanée Chahinian (1) , Aude Allard (2) , Jean-Stéphane Bailly (3) , Katia Chancibault (2) , Fabrice Rodriguez (2) , Christian Salles (1) , Marie-George Tournoud (1) , Carole Delenne (1, 4)

1 HSM - Hydrosciences Montpellier
2 IFSTTAR/GERS/EE - Eau et Environnement
3 UMR LISAH - Laboratoire d'étude des Interactions Sol - Agrosystème - Hydrosystème
4 LEMON - Littoral, Environment: MOdels and Numerics

One of the direct consequences of urban growth is the multiplication of buried utility networks. It is difficult to obtain complete and accurate maps of these networks because they have often been produced by different parties at different times. This work aims at putting forward a methodology to detect manhole covers on high resolution images, in order to reconstruct an urban drainage network. A circular detection filter is applied to high resolution orthophotos and Pleiade images to pinpoint mahole covers. The primary results are encouraging as 42% of manhole covers are detected by this method. Further work is carried out to improve the method and extend it to rectangular grates, that are not detected by the circular filter.

Mots clés: TRAITEMENT DU SIGNAL; IMAGE; TRAITEMENT DES IMAGES; SATELLITE; TELEDETECTION; FILTRE; DETECTION; ASSAINISSEMENT; EGOUT; RESEAU DE DRAINAGE; DRAINAGE; DETECTION D'OBJET; DETECTION DES CONTOURS

Domaines: Sciences de l'ingénieur [physics]; Traitement du signal et de l'image [eess.SP]

Olivier Bartoli, Nanée Chahinian, Aude Allard, Jean-Stéphane Bailly, Katia Chancibault, et al.. Manhole cover detection using a geometrical filter on very high resolution aerial and satellite images. JURSE 2015, Joint Urban Remote Sensing Event, Mar 2015, LAUSANNE, Switzerland. 4p, ⟨10.1109/JURSE.2015.7120521⟩. ⟨hal-01373941⟩