Instance Segmentation for Governmental Inspection of Small Touristic Infrastructure in Beach Zones Using Multispectral High-Resolution WorldView-3 Imagery

Bibliographic Data

ID22033820
AuthorsOsmar Luiz Ferreira de Carvalho (0000-0002-5619-8525, Universidade de Brasília, corresponding author), Rebeca dos Santos de Moura (0000-0002-7685-8826, Universidade de Brasília), Anesmar Olino de Albuquerque (0000-0003-1561-7583, Universidade de Brasília), Pablo Pozzobon de Bem (0000-0003-3868-8704, Universidade de Brasília), Rubens de Castro Pereira (Universidade de Brasília), Li Weigang (0000-0003-1826-1850, Universidade de Brasília), Dibio Leandro Borges (0000-0002-4868-0629, Universidade de Brasília), Renato Fontes Guimarães (0000-0002-9555-043X, Universidade de Brasília), Roberto Arnaldo Trancoso Gomes (0000-0003-4724-4064, Universidade de Brasília), Osmar Abílio Carvalho Júnior (0000-0002-0346-1684, Universidade de Brasília, corresponding author)
Year2021
Volume10
Issue12
Pages813
Publication date2021-11-30
Open AccessYes
TypeARTICLE
VenueISPRS International Journal of Geo-Information (JOURNAL)
Journal identifiersISSN: 2220-9964 • E-ISSN: 2220-9964
PublisherMDPI AG (PUBLISHER • IT)
DOI10.3390/ijgi10120813
OpenAlexW3215280598
LanguageEN
References cited91

Misappropriation of public lands is an ongoing government concern. In Brazil, the beach zone is public property, but many private establishments use it for economic purposes, requiring constant inspection. Among the undue targets, the individual mapping of straw beach umbrellas (SBUs) attached to the sand is a great challenge due to their small size, high presence, and agglutinated appearance. This study aims to automatically detect and count SBUs on public beaches using high-resolution images and instance segmentation, obtaining pixel-wise semantic information and individual object detection. This study is the first instance segmentation application on coastal areas and the first using WorldView-3 (WV-3) images. We used the Mask-RCNN with some modifications: (a) multispectral input for the WorldView3 imagery (eight channels), (b) improved the sliding window algorithm for large image classification, and (c) comparison of different image resizing ratios to improve small object detection since the SBUs are small objects (

Computer vision · Multispectral image · Pixel · Remote sensing · Segmentation · Sliding window protocol · Automated Road and Building Extraction · Computer Science · Geochemistry and Geologic Mapping · Remote-Sensing Image Classification · Artificial Intelligence · Geology

Tools

Open DOIOpen Access