Skip to content
Projects
Groups
Snippets
Help
Loading...
Help
Submit feedback
Contribute to GitLab
Sign in
Toggle navigation
I
imagej-elphel
Project
Project
Details
Activity
Releases
Cycle Analytics
Repository
Repository
Files
Commits
Branches
Tags
Contributors
Graph
Compare
Charts
Issues
3
Issues
3
List
Board
Labels
Milestones
Wiki
Wiki
Members
Members
Collapse sidebar
Close sidebar
Activity
Graph
Charts
Create a new issue
Commits
Issue Boards
Open sidebar
Elphel
imagej-elphel
Commits
6da5747a
Commit
6da5747a
authored
Feb 25, 2026
by
Andrey Filippov
Browse files
Options
Browse Files
Download
Email Patches
Plain Diff
Moved project docs to the git-controlled directory
parent
20be6135
Changes
2
Expand all
Hide whitespace changes
Inline
Side-by-side
Showing
2 changed files
with
349 additions
and
0 deletions
+349
-0
project-details.md
docs/project-details.md
+301
-0
project-summary.md
docs/project-summary.md
+48
-0
No files found.
docs/project-details.md
0 → 100644
View file @
6da5747a
This diff is collapsed.
Click to expand it.
docs/project-summary.md
0 → 100644
View file @
6da5747a
# Project Summary (imagej-elphel)
## Context
-
Java ImageJ 1.x plugin evolved over ~10 years; code is large and entangled due to shifting applications.
-
Original target: vehicle navigation in complete darkness (DARPA “Invisible Headlights”).
-
Later applications: drone‑based mine detection, long‑range drone detection, and most recently foliage penetration from a Cessna.
-
Hardware is now stable: 16‑sensor LWIR camera; processing runs in batch mode with list files.
## Current workflow
-
Main batch entry path:
`Eyesis_Correction.runMenuCommand()`
→
`buildSeries()`
→
`TwoQuadCLT.buildSeriesTQ()`
→
`OpticalFlow.buildSeries()`
.
-
Batch inputs are
`.list`
files such as:
`/media/elphel/btrfs-data/lwir16-proc/NC/lists/nc_site_32A.list`
Example line:
`1763232778_911052 1, -2, -1, 0, 0`
(start at scene 1, skip last 2; other params unused here)
-
Scenes live under:
`/media/elphel/btrfs-data/lwir16-proc/NC/scenes/`
(timestamp folders; “_” used instead of “.”)
## Output data
-
Batch processing produces many per‑reference outputs (images, XML, CSV) per 8‑second sequence.
-
Calibration files of interest:
`*-FIELD_CALIBRATION.corr-xml`
Example:
`/home/elphel/lwir16-proc/NC/models/1763232117-1763234145/1763232535_880101/1763232535_880101-FIELD_CALIBRATION.corr-xml`
-
These contain per‑sensor azimuth/tilt/roll corrections (radians), e.g.:
`EYESIS_DCT_AUX.extrinsic_corr_azimuth9`
,
`EYESIS_DCT_AUX.extrinsic_corr_roll8`
, etc.
## Immediate goals
1) Scan data directories for anomalies (missing files, bad CSVs, etc.).
2) Aggregate calibration data across sequences (LY “lazy eye”):
-
Extract azimuth/tilt/roll per sensor from
`*-FIELD_CALIBRATION.corr-xml`
.
-
Build CSV for Calc to visualize sensor drift (notably sensors 3 and 8).
3) Prepare data for university DNN training and scene classification.
4) Generate map screenshots from GPS data (later; network/GUI dependent).
## Global LMA pause/resume note
-
Current Global LMA/quarter-generation parked state and next-step backlog are documented in:
`attic/CODEX/HOWTOs/project-details.md`
under
`Parked State Snapshot (2026-02-25)`
.
## MCP thoughts (high‑level)
-
MCP should be thin and flexible, with a small set of batch tools and a generic parameter map.
-
Avoid GUI automation; use batch‑mode entry points and Properties‑based parameter IO.
-
A minimal MVP is feasible: one batch command + parameter map + status/logs.
## Logging
-
Added
`LogTee`
helper (in repo) to tee
`System.out/err`
to per‑scene files without changing existing printlns.
## Notes
-
The code is expected to change; tooling should allow rapid iteration with minimal refactor.
Write
Preview
Markdown
is supported
0%
Try again
or
attach a new file
Attach a file
Cancel
You are about to add
0
people
to the discussion. Proceed with caution.
Finish editing this message first!
Cancel
Please
register
or
sign in
to comment