This presentation introduces a new step-by-step approach that aims at helping to fix the incompleteness and heterogeneity of the OSM data.
It first focuses on the OSM features that comprise the backbone of homogeneous OpenStreetMap data, before broadening the perspective to include other objects: the nodes representing settlements, the main road connecting them to the main road network, as well as the residential area showing the extent of the settlement.
The OSM Skeleton project aims to identify all missing, inaccurate, incomplete, or obsolete objects in the OSM data backbone in order to resolve inconsistencies. OSM Skeleton uses external data to identify the missing features and performs spatial analysis, crossed features and various metrics to analyze incompleteness and heterogeneity. For instance, the approach introduces a hybrid analytic feature that combines residential and street network areas.
Each type of issue, whether confirmed or potential, is represented by a specific layer at national scale and can be resolved by clicking to access the relevant area in an OSM editor to modify the existing OSM data using the available imagery. The most significant issues are displayed at all scales, while minor ones are displayed at higher zoom levels.
All the issues can be processed by a click giving access to a webpage providing mapping instructions and buttons to open the area in either JOSM or iD. Forecasting next mapping steps involving field mapping, the OSM Skeleton approach also performs an analysis of the completeness of points of interest (POIs) for cities and towns.
Initially developed in two regions of Africa and South America, the methodology reveals regional or national patterns, highlighting significant contrasts between regions and between small towns with the same region. Daily updated statistics categorized by issue or first-level administrative division provide a complete overview of the issues at hand.
Technically, the OSM Skeleton approach uses standard PostgreSQL-generated Open Geospatial Consortium (OGC) layers. These layers can be used on ready online maps provided by a spatial data infrastructure (SDI) or in any OGC-compatible client. All layers can be downloaded natively in various GIS formats. Rather than relying on remote mapping, OSM Skeleton aims to expand local OSM activities, including imagery and field mapping, as well as the local promotion of enhanced, complete OSM data.