Detection of House Constructions by Their GPS Location. Case Study: Rural Areas of Madagascar
Domaine:
geospatialType de record:
datasetpaper
Créateur:
Jos
Éditeur:
Ins
Hôte:
In Madagascar, traditional houses are usually built with red raw earth or wood for the walls, and thatched roofs. Their rectangular shape, oriented North-South, represents an architectural identity that is now threatened with disappearance. Mapping these traditional constructions requires their identification and geographical localization. In this study, we propose an approach based on computer vision and satellite image analysis. We created a new dataset to train the YOLOv5 OBB algorithm, applied to the detection of traditional houses. The application image is divided into sub-images, results are reorganized within the global image, and filtered according to the North-South orientation. Satellite image projection then makes it possible to determine the geographical coordinates of these constructions, assuming the availability of a known reference point. Our method thus highlights the joint consideration of orientation and roofing as the main criteria for identifying and mapping traditional Malagasy houses.
Similaires
The location of the hazard model study areas within Eden.Enhancing Fault Detection and Location Identification using GIS-based Techniques: A Case Study of Arba Minch Distribution SystemMachine Learning for Fault Detection, Classification, and Location in Transmission Lines: A Case Study of Bonga–Mizan, EthiopiaReal-time Detection of Evoked Potentials by Deep Learning: a Case StudyA Deep Learning Classification Approach Using High Spatial Satellite Images for Detection of Built-Up Areas in Rural Zones: Case Study of Souss-Massa Region, MoroccoWhy socio-economic and attitudinal factors cannot predict entomophagy in rural areas of Madagascar