Extraction of Parcel Boundary from UAV Images Using Deep Learning Techniques

Khadanga, Ganesh; Jain, Kamal

doi:10.1007/978-981-16-0425-6_12

Cited by 1 publication

(1 citation statement)

References 8 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Previous work has investigated the automatic detection of boundaries from aerial and satellite imagery. These boundaries can be either from buildings [26][27][28][29][30][31] or cadastral boundaries [32][33][34][35][36][37][38]. Since the focus of this work is on the automatic extraction of cadastral boundaries, we will now briefly review previous work touching on this topic.…”

Section: Detection Of Boundaries Using Machine Learning and Deep Lear...mentioning

confidence: 99%

The SmartLandMaps Approach for Participatory Land Rights Mapping

Lindner,

Degbelo,

Vassányi

et al. 2023

Land

View full text Add to dashboard Cite

Millions of formal and informal land rights are still undocumented worldwide and there is a need for scalable techniques to facilitate that documentation. In this context, sketch mapping based on printed high-resolution satellite or aerial imagery is being promoted as a fit-for-purpose land administration method and can be seen as a promising way to collect cadastral and land use information with the community in a rapid and cost-effective manner. The main disadvantage of paper-based mapping is the need for digitization to facilitate the integration with existing land administration information systems and the sustainable use of the data. Currently, this digitization is mostly done manually, which is time-consuming and error-prone. This article presents the SmartLandMaps approach to land rights mapping and digitization to address this gap. The recording involves the use of sketches during participatory mapping activities to delineate parcel boundaries, and the use of mobile phones to collect attribute information about spatial units and land rights holders. The digitization involves the use of photogrammetric techniques to derive a digital representation from the annotated paper maps, and the use of computer vision techniques to automate the extraction of parcel boundaries and stickers from raster maps. The approach was deployed in four scenarios across Africa, revealing its simplicity, versatility, efficiency, and cost-effectiveness. It can be regarded as a scalable alternative to traditional paper-based participatory land rights mapping.

show abstract