Photogrammetry DTM reconstruction performed with the JANUS LEGA data
Benoît Seignovert
Simeon Schmauß
Léo Scordia
Marion Massé
Stéphane Le Mouélic
Gabriel Tobie
Laetitia Le Deit
Janus STM - Sept. 2026 - London
Photogrammetry vs. regular stereogrammetry
Very popular with
drone photography
2
2
3
4
3
2..n
©️ Wingtra
Beyer et al. 2018
Photogrammetry
expertize in Nantes (LPG)
Le Mouélic et al. 2020
Le Mouélic et al. 2026
Halley's Comet
The Bayeux tapestry example
Le Mouélic et al. 2026
Sold out!
The Bayeux tapestry example
Does it work on
JANUS data?
YES!
First attempt in August 2024:
- just after the LEGA flyby
- use JPEG preview images
- uncalibrated data
- no camera model
- no geometric information
- mix of filters
Alignment success on 18 images
👉 1st JANUS DTM!
21:24:30
21:24:01
…
How it works
Flight
path
| time | 21:24:27 | 21:24:37 |
|---|---|---|
| emi | 52° | 55° |
| alt | 1605 km | 1630 km |
| res | 49 m/pix | 54 m/pix |
Global alignment 👉 tie points
Images aligned
Tie points
Initial input:
- SPICE kernels location and expected attitude
- sensor size and expected focal length
- camera profile unconstrained
LRO basemap
36 images aligned:
- 14 PAN
- 3 R / G / B / V / CMT med / MT strong
- 2 NIR 1 & 2
👔 20k tie points identified
LRO basemap
⚙️ Mesh from dense points cloud
LRO basemap
📍 Ortho image
+ reference points to constrain the geographic coordinates on the Moon reference frame (ME)
LRO basemap
Output JANUS DTM
Comparison with Kaguya/LOLA
Kaguya resolution: 59 m/pix | 3-4 m accuracy
Longitude
profiles
1
2
3
Results
Results
Δh = -10 ± 220 m
(Janus pixel scale: 50m/pix)
Latitude
profiles
4
5
6
Results
Δh = -90 ± 150 m
(Janus pixel scale: 50m/pix)
Strengths 💪
Very low parallax requires
non-nadir looking
Not as accurate as
dedicated stereo pairs
Possible skew in GCO
due to rolling shutter
(can't be tested during the Cruise)
Opportunistic acquisitions
no additional DV
Single pass acquisition
Low impact on global coverage
Can be used with multi-filters
sequences (eg. R/G/B)
Fixed pointing
works on flyby ingress/egress or
limb to nadir looking slews
📣 Photogrammetry is not a replacement for stereo observations!
Weaknesses 🏋
EGA 2
To be tested over Madagascar and Africa.
| emi | 40° | 60° |
|---|---|---|
| lon | 50°E | 28°E |
| alt | 8700 km | 8900 km |
| res | 200 m/pix | 350 m/pix |
| Δh | 0 km | 2.5 km |
EGA 1 - Cloud tops
No ground truth
Relevant for WG4 on Jupiter?
Other opportunities with NavCam (?)
LGA (2024) - not implemented
EGA 2 (2026) - not implemented
2024 - Not implemented by the SOC/MOC
2026 - Not implemented in favor of JANUS/MAJIS/NavCam synergy
2029 - Possible but major burst-mode limitations
Other opportunities with JMC
2024 - Yes! on the Moon
2026 - Planned for EGA2
2029 - TBC
microcameras.space
40 images aligned
x2
Global shape modeling