Kristof Ostir

University of Ljubljana, Faculty of Civil and Geodetic Engineering, Faculty Member

Research Centre of the Slovenian Academy of Sciences and Arts, Institute of Anthropological and Spatial Studies, Faculty Member

Followers

Following

Co-authors

Public Views

University of Ljubljana

Tomislav Popit

Anka Lisec

University of Ljubljana

Tu Wien

University of Ljubljana

Interests

Uploads

Papers by Kristof Ostir

Delineation of vacant building land using orthophoto and lidar data object classification

Geodetski vestnik, 2019

In the latter, sustainable spatial development is defined as development providing for the needs ... more In the latter, sustainable spatial development is defined as development providing for the needs of current and future generations, taking into account the objectives of spatial planning in a way to allow for rational and efficient land use and the use of buildings, and coordinated arrangement of activities in space (Article 6 of ZUreP-2, 2017). Following the strategic spatial guidelines, spatial planning is embodied in implementing planning acts, i.e. on building plots intended for construction (Article 3 of ZUreP-2, 2017). Building land can be built-up (ground occupied by buildings) or vacant (Article 142 of ZUreP-2, 2017). Vacant building land, i.e. land where construction may take place, is relevant for spatial development. The act provides for built-up and vacant land to be entered in the Building Land Registry, meaning that this information should be acquired and serve as a basis for spatial analyses, spatial documents, and the planning of fitting-out of space and land policy measures. Currently there is no appropriate record of built-up and vacant land in Slovenia (Medved and Berk, 2019), while the municipalities that need this information the most have addressed this issue in various ways. This is being tackled through mass collection of data on inhabited land and the actual use of inhabited land (Masovni zajem …, 2017), which was started in 2017. Now, 24% of all built-up plots in Slovenia has been recorded (Medved and Berk, 2019) and according to the work programme all land across Slovenia should be logged by the end of 2020. The website http://storitve.pis.gov.si/pis-jv/evidenca_stavbnih_zemljisc.html provides access to all recorded data for Prekmurje, southeast Slovenia, and the municipalities of Piran and Izola. Data are acquired (Masovni zajem …, 2017) by photointerpretation, which is supported by other data sources. The acquisition and maintenance (updating) of spatial data using the visual photointerpretation method is a demanding task, financially unfavourable, and time-consuming (Kos and Šumrada, 2013; Mesner, 2016). Remote sensing procedures are used to ensure that the data are being updated. Object-oriented analysis is often used to identify the changes in the Earth's surface, land management, and planning (Kanjir, 2009; Mesner, 2016). The breakthrough of object-oriented analysis is attributed to the development of satellite systems reaching a spatial resolution better than 1 m (Veljanovski et al., 2011). The data acquired are useful in geographic information systems (GIS), in general and, also for the needs of spatial planning. Object-based classification is a procedure where pixels with similar characteristics are combined into segments. In the semantic classification procedure, the segments are, based on their attributes, classified into semantic classes (Liu and Xia, 2010; Veljanovski et al., 2011). As underlined by Liu and Xia (2010), Blaschke (2010), and Yanjing (2015), segmentation should be given special attention during object-based classification. A well-designed segmentation maintains the right relationship between the size of the segments and the size of the structures on the ground, so that the segments are neither too fragmented nor too big (Kanjir, 2009; Blaschke, 2010). The segmentation and semantic classification phases are followed by post-classification, which corrects obvious errors and allows for generalisation of the results (Kokalj and Oštir, 2006; Veljanovski et al., 2011).

Kristof Ostir

Related Authors

Uploads

Papers by Kristof Ostir

Log In