ternaus/TernausNetV2
TernausNetV2: Fully Convolutional Network for Instance Segmentation
This helps urban planners, geospatial analysts, and city developers automatically identify and map building footprints from satellite imagery. It takes high-resolution satellite photos (RGB, panchromatic, and multi-spectral) and outputs precise, pixel-level outlines of individual buildings. This is for professionals who need accurate building data for large-scale urban analysis.
546 stars. No commits in the last 6 months.
Use this if you need to precisely detect and delineate individual building outlines from satellite images for urban planning, monitoring, or mapping tasks.
Not ideal if you're looking to classify broad land use types or analyze features other than building footprints.
Stars
546
Forks
112
Language
Jupyter Notebook
License
BSD-3-Clause
Category
Last pushed
May 23, 2020
Commits (30d)
0
Get this data via API
curl "https://pt-edge.onrender.com/api/v1/quality/computer-vision/ternaus/TernausNetV2"
Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.
Related tools
MarvinTeichmann/MultiNet
Real-time Joint Semantic Reasoning for Autonomous Driving
WangLibo1995/GeoSeg
UNetFormer: A UNet-like transformer for efficient semantic segmentation of remote sensing urban...
cvjena/cn24
Convolutional (Patch) Networks for Semantic Segmentation
rankseg/rankseg
Boost segmentation model mIoU/Dice instantly WITHOUT retraining. A plug-and-play, training-free...
arplaboratory/STHN
[RA-L 2024] Official repository for "STHN: Deep Homography Estimation for UAV Thermal...