Distance to neighbour calculations among OBIA primitives as an innovation to urban mapping techniques. (2nd January 2018)

• Record Type: Journal Article
• Authors: de Kok, R.; Wężyk, P.; Hejmanowska, B.; Książek, J.
• Is Part Of: International journal of image and data fusion. Volume 9:Number 1(2018)
• Page Start: 21
• Page End: 42
• Publication Date: 2018-01-02
• Subjects: Urban Atlas -- urban fabric -- big data -- distance to neighbour -- perception -- OBIA; Image processing -- Periodicals; Multisensor data fusion -- Periodicals; Multisensor data fusion; Periodicals; 621.36705
• DOI: 10.1080/19479832.2017.1375029
• Languages: English
• ISSNs: 1947-9832

Abstract

ABSTRACT: The intuitive notion of urban fabric is related to the type of buildings and their proximity. Mapping the density of urban fabric by proxy, using the vegetation index, is considered standard. The classification of urban fabric can also be achieved using the information of all individual buildings. By using the volume and measuring their distance to all other image objects in the map, a new urban density map is made. This requires all building volumes in the city and fusing such data with remote-sensing imagery. The distance calculation of over 1.3 million segments in 6 categories, results in a new map (of Cracow) with urban density and is compared to the existing urban fabric map. The Normalized Difference Vegetation Index and the commercial off the shelf (COTS) roadmap are traditionally recommended for urban mapping. But COTS data can, at this stage, not be automated. The mass processing of the distance to categorical buildings, with object-based image analysis (OBIA), can be automated instead. The OBIA classification requires image–object fusion for the transfer of attributes in various levels of aggregation. This procedure can be applied in extending the manuals and it serves as an innovation for the standard urban map procedure.