# ethz-asl/segmap

A map representation based on 3D segments

Repository: https://github.com/ethz-asl/segmap
Canonical: https://ross.abutalabs.com/products/segmap
Language: C++
License: BSD-3-Clause
License Family: permissive
Last push: 2021-03-17T22:30:59+00:00

## Health v2 (maintenance only)
Score: 32/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 0, release rhythm 35, longevity 100
- inputs: {"age_days": 3634, "days_push": 1995, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: no_releases
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 1096, forks 389 (observed 2026-08-28T04:03:34.333004+00:00)

## What it is
SegMap is a C++ library and ROS package providing a segment-based 3D map representation for robot localization, environment reconstruction, and semantic extraction from LiDAR point clouds. It includes a 3D CNN encoder-decoder for data-driven segment descriptors, LiDAR odometry, and single/multi-robot SLAM back-ends.

## Use cases
- build a 3D segment-based map from LiDAR data for a mobile robot
- localize a robot within a previously mapped 3D environment
- perform place recognition in 3D point clouds
- extract semantic information from segmented LiDAR scans
- run multi-robot SLAM with shared segment-based maps
- train and evaluate neural network descriptors for 3D segments

## When to choose
- you need segment-based mapping and localization for LiDAR-equipped robots
- you want research-grade SLAM with learned 3D segment descriptors
- your project already uses ROS and catkin on Ubuntu

## When to avoid
- you need a production-supported, actively maintained SLAM stack
- you work outside ROS/Linux or with camera-only data
- you need a simple drop-in library without deep learning dependencies like TensorFlow

## Facets
- artifact type: library
- maturity: maintenance
- function: machine-learning, computer-vision, simulation, graphics, sdk
- domain: robotics, autonomous-vehicles, machine-learning, computer-vision
- platform: cpp, python
- tags: slam, lidar, point-cloud, segmentation, place-recognition, localization, mapping, ros, 3d-cnn, research-code, algorithms, linux

## Member repositories
- ethz-asl/segmap (main) score 32

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:03:34.333004+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-30T06:47:10.559856+00:00, confidence not recorded.
  - readme: https://github.com/ethz-asl/segmap (fetched 2026-08-28T04:03:34.333004+00:00, sha 5ec61929bcd0)
- Data as of 2026-08-30T08:39:29.467469+00:00.