AI-Based Automated Urban Parcel Mapping and Cadastral Feature Extraction System using Drone lmagery
Ministry of Rural Development · Robotics and Drones · Software
The rare land-records statement with genuinely available training data and a demo that explains itself, but the differentiator is clean parcel topology rather than segmentation accuracy, so budget most of your time downstream of the model.
What it actually is
Turning drone photographs of a town into a property map is currently done by someone tracing boundaries by hand, which takes months. The ask is software that looks at the drone imagery and draws the plot boundaries, buildings and roads automatically. A human surveyor should then be able to correct the result in a web map.
What to build
An automated cadastral extraction pipeline taking orthorectified drone imagery with DSM and DTM as input, running segmentation models to produce four output classes the description names — parcel boundaries, building footprints, roads and access corridors, and land-use classification — then vectorising those masks into topologically valid parcel polygons with shared edges rather than overlapping blobs, running a validation pass that flags overlapping or inconsistent geometries and gaps against existing GIS parcel layers, and delivering it through a Web-GIS interface where a surveyor reviews each extracted parcel, edits vertices, accepts or rejects, and exports GIS-ready cadastral output.
Smallest thing that wins the room
Drop a raw drone orthophoto tile of a dense settlement onto the map and watch parcel polygons and building footprints appear over it in seconds, then drag a vertex and show the shared boundary of the neighbouring parcel move with it.
How crowded this one gets
A guess, projected from the 2025 statements — the last year where both the submission counts and the winners were published.
Quieter than 60% of the 226 · #91 of 226 by expected field
A normal-sized field. Your idea has to be good, not miraculous.
Why: central ministry statements sat below the average.
This is a guess, not a fact
Nobody has published 2026’s numbers yet. This is an analysed estimate from last year’s pattern, so please do not take it as the truth — check the live counter on the SIH portal before you decide anything. The range covers the middle half of likely outcomes, so one statement in two lands outside it. Entry closes at 500 ideas per statement, so no range goes past that — a statement that reaches the cap fills and shuts rather than drawing an unlimited crowd. The model reads only three things a team can see before choosing — software or hardware, the theme, and what kind of body posted it — and those explain about a quarter of the variation in last year’s field sizes (R² 0.25 on held-out statements). Trust the band more than the number, and the ordering more than either. It cannot see how good your idea is, which is the part that actually decides it.
The scores
The number is the shorthand. The line under it is the reason.
Acceptance potential
4/5Real public datasets remove the data risk entirely, the deliverable is bounded and clearly specified, and the demo is visual — the main drag is that building footprint extraction is a well-worn CV task, so your differentiation has to come from the cadastral topology work rather than the segmentation.
Feasibility
4/5This is one of the best-supported computer vision tasks in the open data world — INRIA Aerial Image Labeling, SpaceNet, Open Cities AI and OpenAerialMap all provide labelled aerial imagery, Microsoft and Google have released building footprints for India, and segmentation architectures for this exact task are mature and pretrained.
Innovation scope
2/5Five numbered platform components fix the entire pipeline from segmentation through topology generation to the Web-GIS editor, and the input datasets and output classes are both enumerated, so you are implementing a specified architecture rather than designing one.
Clarity
4/5The inputs, the four feature classes to extract, the five platform components and the expected outputs are all stated explicitly, and the only real gap is that no accuracy threshold or positional tolerance is given for what counts as a correctly extracted boundary.
Effort
HeavyModel training is the easy part; raster-to-vector polygonisation that produces clean shared edges, automated topology repair and a usable Web-GIS editing interface are each harder and slower than the segmentation, and teams consistently discover this late.
Demo-ability
EasyWatching boundaries draw themselves over a photograph of a real neighbourhood is immediately legible to anyone in the room and needs no explanation of what the model is doing.
In its favour
- Green flag: Multiple large labelled aerial datasets are public, so unlike almost every other statement in this ministry block your project cannot collapse for lack of data
- Green flag: The DSM minus DTM height difference gives you a building mask almost for free, which is a strong classical baseline you can ship even if the deep model underperforms
- Green flag: The four output classes are named in the description, so your scope is fixed by the statement and cannot be criticised as conveniently chosen
- Green flag: The self-drawing boundary demo lands with judges in under ten seconds, which matters more at a hackathon than model architecture
Against it
- Red flag: Building segmentation is a heavily solved benchmark task and a judge may reasonably ask what you added beyond running a pretrained model on a new tile
- Red flag: Parcel boundaries are legal fictions that are frequently invisible in imagery — a fence, a wall or nothing at all — so a model that finds buildings well will still miss the boundaries the statement is actually about
- Red flag: Turning masks into a clean topological parcel layer with shared edges and no slivers is genuinely hard and is where most of your time will go, not the model
- Red flag: Open datasets are dominated by European and American cities; Indian informal settlements with irregular geometry and overlapping roofs look nothing like them, and transfer performance drops sharply
What you will be writing
- SAM / U-Net / Mask R-CNN aerial segmentation
- GDAL + rasterio orthoimagery processing
- shapely + topojson polygonisation with shared edges
- PostGIS topology validation
- GeoServer + OpenLayers Web-GIS editor
- DSM/DTM normalised height differencing
- Geospatial AI
- Urban land records
- Computer vision
Prior art to read before you start
building footprint extraction · automated parcel boundary delineation · Web-GIS cadastral editing
Analysed by Claude Opus. Every score above is a judgment call with its reasoning attached — kindly cross-check this against the official statement on the SIH portal before your team commits to it.