ECO: Incremental Ego-Centric Octree Update for Point Streams
2026-07-06 • Robotics
RoboticsGraphics
AI summaryⓘ
The authors created a new data structure called the Ego-Centric Octree (ECO) to help robots quickly understand their 3D surroundings as they move. Instead of mapping everything globally, ECO focuses only on the robot’s nearby area using a kind of moving window, which helps avoid slow updates and unbalanced maps. Their method updates the map efficiently by sorting parts of the environment into categories and skipping unnecessary work. Tests showed that ECO updates faster and reduces delays in robot tasks compared to older methods. It also helps keep track of moving objects for a short time without slowing down the system.
octreespatial data structure3D mappingincremental updatereal-time processingrobot perceptionvoxel mapdynamic environmentKITTI benchmark
Authors
Jaemin Yu, Seongyoon Jeong, Kang-Wook Chon, Duksu Kim
Abstract
Constructing octrees for mobile robots that process continuous point streams in real time poses significant computational and memory challenges. Standard global structures often suffer from high latency and unbalanced tree growth. We introduce the Ego-Centric Octree (ECO), a spatial data structure that acts as a 3D sliding window, dynamically bounding the mapping space to the robot's immediate surroundings. ECO uses an efficient incremental update algorithm that categorizes the environment into shift-out, shift-in, and overlap regions, eliminating redundant global coordinate transformations. Evaluations on the KITTI benchmark demonstrate that ECO reduces update times by up to 25.60% (24.87% on average) compared to full static reconstruction and by up to 67.52% (54.60% on average) compared to a bounded incremental baseline. Furthermore, ECO substantially lowers the total system latency of downstream tasks, running up to 34.17% faster than full reconstruction in voxel-map generation. In dynamic scenes, ECO naturally retains a short-term temporal memory of moving objects, providing useful temporal context while keeping update cost bounded and the tree balanced for real-time spatial perception.