Street-level imagery analysis in QGIS: sample a street network, match open Panoramax and Mapillary images or your own geotagged photos, segment them locally and map green view index, sky view factor, tree and building view factors and visual enclosure per street segment and area, with an HTML report that names every formula, source and licence.
02StreetView carries a street-level imagery study from a street network to mapped, aggregated indicators in one dock panel.
Sampling: points every few metres along the street lines with the local street bearing, in a projected working CRS (the UTM zone of the extent for geographic layers).
Sources: Panoramax (federated open imagery, no account), Mapillary (your own free client token, stored in the QGIS authentication database) and local folders of geotagged JPEGs. The licence and author of every image are read from the service and kept with it; images stay in a local cache and there is no image export.
Segmentation: built-in colour rules (numpy; sky and vegetation) or the UperNet + ConvNeXt-T ADE20K scene-parsing model (MIT weights, downloaded on demand and checked by SHA-256) through the optional onnxruntime package; custom ONNX models with a JSON class map are accepted. No PyTorch.
Indicators: green view index (solid-angle weighted on 360 degree panoramas, and the pixel share for comparison), sky view factor, tree and building view factors, sky share, visual enclosure and building, road, sidewalk and vehicle shares, each with its formula, unit and reference.
Aggregation: means and medians per street segment and per area with coverage and a low-coverage flag; an optional DSM check compares photo-based SVF with horizon SVF from the DSM at the same points.
Output: layers styled in a 02StreetView group, a self-contained HTML report (formulas typeset with KaTeX, offline), CSV and GeoPackage with attribution and licence fields.
Only qgis.core, GDAL, numpy and PyQt are required; SciPy is used when present and onnxruntime is optional.
Manual: https://geophilo.com/zero2streetview/
Plugin Tags