Bulldozer is an open-source CNES tool that extracts a Digital Terrain Model from a DSM without requiring surface maps or exogenous data by using a cloth simulation method.

Tri-stereo acquisitions with one image near vertical are particularly useful in dense urban areas and mountainous regions where occlusions occur.

xDEM is an open-source software created by a collective of glaciology researchers (GlacioHack) and developed in collaboration with CNES for analyzing Digital Elevation Models.

The CO3D constellation enables stereo-synchronous pairing by observing the ground simultaneously with two satellites from different angles.

3D change detection can reveal partial building damage (for example, lost floors) that 2D change detection methods may miss.

CNES technical teams are developing algorithms to infer 3D models from satellite data such as the CO3D mission.

CARS performs DSM production via four main steps: resampling, matching, triangulation, and rasterization.

Moving objects between temporally separated stereo acquisitions (for example cars, trains, or vegetation in wind) can change position and complicate 3D reconstruction from those images.

Rasterization projects 3D points onto a 2D grid to create a matrix image that assigns an altitude to each pixel, producing the final DSM.

Classical photogrammetry measures pixel displacement (parallax) between two images taken from different angles to determine altitude by triangulation.

Space imagery acquired through mechanisms such as the International Charter Space and Major Disasters can be obtained within hours of an event for rapid damage mapping.

CNES developed the open-source tool PICANTEO for 3D change detection and applied it to study damage from the February 2023 Kahramanmaraş earthquake.

A higher B/H ratio increases measured parallax between images and generally improves altimetric precision, subject to occlusions in urban canyons.

The CO3D mission consists of a constellation whose four satellites were launched on July 25, 2025 from the Guiana Space Center aboard a Vega-C launcher.

CNES teams report that first CO3D data currently being calibrated include a DSM with a 2-meter spatial resolution on the Pic du Midi.

A Digital Surface Model is a 2.5D representation of the Earth's surface where each image pixel (x,y) is associated with an altitude (z).

A low B/H ratio (small baseline relative to altitude) yields low measured parallax and can decrease altimetric precision.

CARS is an open-source software developed by CNES that produces Digital Surface Models from one or more stereoscopic image pairs.

xDEM functionality includes DEM co-registration, terrain attribute calculation (slope, shadow, aspect, curvature, roughness), uncertainty analysis, and bias corrections.

Space BD, based in Chuo, Tokyo, signed a Memorandum of Understanding (MOU) with Kick Space Technologies, a space startup originating from Kyushu Institute of Technology located in Kitakyushu, Fukuoka Prefecture, on January 15.

Space BD will coordinate launch opportunities tailored to the missions and characteristics of the satellites developed by KST.

Space BD will promote collaboration throughout various phases of mission management, allowing KST and its customers to focus on the development of satellites and components.

The collaboration aims to expand the supply chain options for the industry and create sustainable business opportunities for both Space BD and KST.

Space BD plans to leverage its international and domestic customer base to support the sale and export-import assistance of space equipment components developed by KST.

The MOU aims to establish a collaborative relationship between Space BD and KST for satellite mission development and launch opportunities.

xDEM provides DEM co-registration, terrain attribute calculation (slope, shadow, aspect, curvature, roughness), uncertainty analysis, and bias corrections.

Bulldozer is an open-source tool developed by CNES that extracts a Digital Terrain Model (DTM) from a DSM without requiring exogenous data such as water masks.

ICEYE and ICESat-2 are radar (SAR and I-SAR) missions identified as notable for 3D Earth observation.

The CARS processing chain relies on four main steps: resampling, matching, triangulation, and rasterization.

xDEM is an open-source software created by a collective of glaciology researchers (GlacioHack) and developed in collaboration with CNES for analyzing Digital Elevation Models.

Triangulation converts pixel displacements into 3D positions by intersecting lines of sight to produce a 3D point cloud (x,y,z).

CNES uses xDEM to qualify 3D products by comparing satellite-derived DSMs with high-resolution LiDAR DSMs used as reference.

A DSM produced from CO3D data with a displayed spatial resolution of 2 meters was shown over the Pic du Midi in the Pyrenees.

PICANTEO is an open-source tool developed at CNES that demonstrated the added value of 3D change detection for the February 2023 Kahramanmaraş earthquake.

CARS supports input images from Pléiades, WorldView, or any optical satellite allowing stereo acquisitions.

Pandora is software developed at CNES that performs the matching step and generates a displacement (disparity) map highlighting successful and failed matches.

The four satellites of the CO3D mission were launched on July 25, 2025 from the Guiana Space Center aboard the Vega-C launcher.

ICESat and GEDI are referenced as LiDAR missions relevant to 3D Earth observation.

CNES technical teams are developing algorithms to infer 3D models from satellite data such as the CO3D mission.

The CO3D mission consists of a constellation that allows two satellites to observe the ground simultaneously from different angles for stereo-synchronous pairing.

The International Charter Space and Major Disasters enables acquisition of disaster imagery within hours of an event to support rapid damage mapping.

Tri-stereo acquisitions include one image taken near the vertical and are particularly useful in dense urban and mountainous areas to mitigate occlusions.

Pléiades and Pléiades NEO satellites are agile optical satellites capable of acquiring stereoscopic and tri-stereo images.

A Digital Surface Model (DSM) is a 2.5D representation where each image point (x,y) is associated with an altitude (z).

Rasterization projects 3D points onto a 2D grid to create the final DSM by associating an altitude with each pixel.

CO3D's planned spatial resolution is 1 meter.

Resampling transforms images into epipolar geometry so corresponding pixels lie along the same line to simplify matching.

CARS is an open-source software developed by CNES that produces Digital Surface Models (DSMs) from one or more stereoscopic image pairs.

Bulldozer derives the DTM from a DSM using a cloth-simulation approach that cleans and inverts the DSM, drops a virtual cloth with tension, iterates, and then inverts the cloth to obtain the ground surface.

CARS can be launched via a command line by specifying stereoscopic image pair(s) through a configuration file.