After computing per-scene pXpYD arrays, saves a tile-resolution
hyperstack <ts>-NADIR-DISPARITY-MAPS.tiff with:
  - frames = scenes (labelled by timestamp)
  - slices = pX, pY, D (3 channels)
  - NaN where scene has no pose or tile has no data

Useful for verifying disparity transform coverage before feeding
the rendered images into OpenMVS/COLMAP.
Co-authored-by: Claude <claude@elphel.com>

New methods in OpticalFlow.java:
- getNadirDisparityMaps(): transforms SfM terrain disparity from the
  segment reference into each scene's own camera frame using
  Cuas.transformFromVirtual(). Returns double[][][nscene][ntiles][3].
- renderNadirSequence(): renders each scene from its own position
  (scene_xyz/atr=ZERO3, ref_scene=scene itself) using per-scene
  pXpYD from getNadirDisparityMaps(). ERS rolling-shutter velocity
  corrections are still applied. Output: <ts>-NADIR-MERGED.tiff.

Unlike renderSceneSequence() (TERRAIN-DISP), which warps all frames
to the reference for zero ground parallax, renderNadirSequence() produces
natural-parallax perspective images from each scene's own viewpoint —
the correct input format for OpenMVS/COLMAP dense reconstruction.

New parameter in IntersceneMatchParameters:
- nadir_gen (boolean, default false): checkbox under 'Flat ground' tab,
  before fgnd_gen_optho. Enables nadir rendering in the existing
  TERRAIN-DISP pipeline block.
Co-authored-by: Claude <claude@elphel.com>

Adds filteredResources to the seed template:
- regex filter for node_modules/.git/Java LS artifacts (existing on .137)
- multiFilter to hide attic/ from Eclipse project explorer and build

Copy this to .project on a new machine setup.
Co-authored-by: Claude <claude@elphel.com>

- scripts/lwir_map/lwir_scene_summary.py  (was fopen_nc_summary.py)
- scripts/lwir_map/lwir_map_overlay.py    (was fopen_overlay.py)
- scripts/lwir_map/README.md              (pipeline doc + known limitations)
- Updated catalog.json and scripts/README.md with new names/paths

Scripts read imagej-elphel linked LWIR output and generate per-scene
TSV + GeoJSON/KML/SVG map overlays for satellite atlas navigation.
Not FOPEN-specific; usable for any airplane/drone LWIR acquisition.
Co-authored-by: Claude <claude@elphel.com>

Documents TERRAIN-DISP-MERGED TIFF format, INTERFRAME.corr-xml pose
convention, Elphel→COLMAP coordinate conversion, disparity-to-depth
formula, and the foliage CLI pipeline (no C++ library linking).
Co-authored-by: Claude <claude@elphel.com>

Replaces hardcoded `boolean reuse_non_centered = true` with a config parameter.
Adds a companion `cuas_reuse_centered` parameter: when true, attempts to load
ROUND_TWO_GOOD before starting computation and returns early if found, bypassing
both phases. Both default to true — missing files behave as if set to false.
Wired into dialog, properties (.corr-xml), and copy() in IntersceneMatchParameters.
Co-authored-by: Claude <claude@elphel.com>

Spawns attic/elphel-agent-tools/bin/cpu_profile_hold.py on startup,
which calls power-profiles-daemon HoldProfile('balanced') via DBus.
Released automatically via SIGTERM on JVM shutdown.

Design: KDE AC default is power-saver; this hold raises the profile to
balanced (hardware-driven boost under load) while ImageJ is running,
then releases it so the machine drops back to power-saver on exit.
Co-authored-by: Claude <claude@elphel.com>

Pure-Java prototype CuasRTDetector + RtParams parameters class +
CuasRTLib JNA stub (activate when JNA added to pom.xml).
No dependencies added; build is clean. See design discussion in
attic/imagej-elphel-internal/handoffs/2026-05-21_RealTimeDesign.md.
Co-authored-by: Claude <claude@elphel.com>

corr_offset_eff is now computed after corr_pairs_ref so the upper bound
(corr_pairs_ref / corr_offset) is available; lower bound kept at 1.
Clamping is flagged in the existing println when it activates.
Co-authored-by: Claude <claude@elphel.com>

scaling

05/11/2026: restore pre-fix amplitude (window fix doubled it)

The previous sed replacement of 'tMTF_Bayer.java' matched across the '\t'
tab escape in a string literal, producing illegal '\M' escape. Restored \t.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

Copy-paste errors left the wrong filename in the "X.java is free
software" line. Fixed with a grep scan across all source files.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

DoubleFHT.java had a copy-paste error in its GPL header (wrong description
and wrong filename from MatchSimulatedPattern). Fixed both and added
attribution to the ImageJ FHT.java source (Wayne Rasband, NIH, public domain).

DoubleGaussianBlur.java already had an in-class attribution comment; added
a matching note in the GPL header for consistency.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

All Elphel-authored files in the cuas, ims, mcp, gpu, orthomosaic,
tileprocessor, vegetation, calibration, cameras, common, correction,
dct, readers, and x3d packages. Dates derived from git log (first and
last commit year); mtimes restored with git-restore-mtime before
generating copyright year ranges.

Third-party files excluded: epfl/Differentials_.java (EPFL code),
gpu/JCuda_ImageJ_Example_Plugin.java (JCuda example, (c) Marco Hutter).
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

These four constants described the layout of TDCorrTile.getMismatchVector()
return values and had no business being class-level statics in CuasMotion.
Moved to local scope in the two methods that use them: getTargetsFromCorr2d
and refineMotionVectors.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

Dead method/constant removal (21 methods across 5 files):
- CuasMotion: addMotionVectors, clearFailByStrongerNeighbor, cloneTargetsSequence,
  filterByHorizon, filterMotionScans, findMatchingTarget, getKeyFrameVelocityScale,
  getTargetImages, getTargetsFromHyperAugment_2025, getVectorFieldHyper, showPxPyDs,
  subtractTargetSequence, trimSuffix; constants INDX_SPEED, INDX_CONFIDENCE, VF_TOP_TITLES
- CorrectionFPN: calculateFPN_old
- CuasCenterLma: getAverageRoll
- CuasMotionLMA: applyWindowToData, copyMotion, getRR0, getFullParametersVector
- CuasTile: setParentIndex, getParentIndex (were already commented out)

targetsAnalyzeMCP fixes:
- Correct PAR constants (add PAR_PXYAR=-2, shift others)
- params array now accepts strings ("pxy","pxyar","vpxy","upxy", or LMA_TITLES names)
  in addition to integer indices; default remains ["pxy"]
- Add missing PAR_PXYAR case (px, py, a, rmse, r)
- Add amplitude to PAR_PXY case
- Default case uses LMA_TITLES[npar] key instead of String.valueOf(npar)
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

- Extract precorr_ra, accum_ra, cuas_gaussian_ra in locateAndFreezeTargetsMulti
- Compute fpixels_accum (LPF with cuas_accum_ra) for shiftAndRenderAccumulate in both loops
- Compute fpixels_refine (LPF with cuas_precorr_ra) for refineMotionVectors
- Both use Gaussian or running-average depending on cuas_gaussian_ra flag
- When ra==1, LPF is a no-op (passes fpixels_tum through unchanged)
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

corr_partial_dbg() NPEs on null layers when sequences are skipped
(has_centered=false or no valid frame pairs). Fill null slots with
empty double[num_tiles][] before calling the visualizer.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

- Add FAIL_PENDING=-1 sentinel constant to CuasMotionLMA
- Add initFromMotionSequence(): clones motion entries into target_sequence_multi as sentinels
  (RSLT_FAIL=FAIL_PENDING, RSLT_CENTERED=CENTERED_NO, RSLT_QSCORE=initial RSLT_MSCORE)
- Add syncMotionScoreToSentinels(): keeps sentinel QSCORE current after each getEffectiveStrengthMV
- Decouple use_motion/select_new flags in filter5Targets from motion_sequence==null
- Add motion==null paths in filter5Targets: prohibit setup and is-max ranking for FAIL_PENDING sentinels
- Extend filter5Targets marking step to cover FAIL_PENDING sentinels (use_motion=true, motion==null)
- Update extendMotionScan center-detection to accept FAIL_PENDING as valid center state
- First loop in locateAndFreezeTargetsMulti: call initFromMotionSequence, syncMotionScoreToSentinels;
  pass null to filter5Targets; use applyFilter(target_sequence_multi); use getRemain(target_sequence_multi)
- getRemain(target_sequence_multi): count FAIL_PENDING+CENTERED_NO in num_noncentered
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

Line 8734 was: select_new = (motion_sequence != null) ? select_new_in : false
This caused pass 2 (filter5(null, tsm, false, true, ...)) to enter the
!select_new path — finding best existing good tiles instead of untested
non-centered tiles — so nothing was ever selected and the centered-mode
loop was a no-op.

Fix: decouple select_new from motion_sequence null-check; only use_motion
is forced false when motion_sequence is null.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

operational because of filter5Targets().

Add targets_mv / frame_center_mv / corr_offset_mv overloads to
correlatePairs() and correlatePair().  When targets_mv is non-null,
GpuQuad.setRectilinearMovingTasks() generates per-tile TpTask centers
that follow the target at each frame (cx0 + vx*offset_scale), so the
GPU correlates tiles centered on the moving object and the result is the
residual velocity rather than the full VX displacement.

Callers that don't need this pass through the old uniform-grid overload
unchanged.  Also reverts the wrong "fixed-destination / shifted-read"
masking edit from the previous session; the masking was already correct
(mask tracks the target per frame) before that edit.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

- Masked frames: replace single-frame display with full fmin..fmax stack
  (scene_titles used as slice labels; scroll to watch target stay in place)
- Source frame display removed (not needed alongside masked stack)
- Single-tile corr2d block removed; replaced by full CORR2D multi-slice image
  after the nseq loop — one slice per keyframe, same tiled layout as *CORR2D.tiff,
  slice-labeled from scene_titles, uses corr_border_contrast from CLTParameters
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

- Pre-compute integer-pixel raised-cosine mask kernel once before nseq loop
- Allocate fpixels_masked[nframes][width*height] once; clear per nseq with Arrays.fill
- Restructure masking: parallel outer loop over frames (nFr via AtomicInteger),
  inner loop over tiles — no write contention (non-overlapping tile guarantee)
- Pre-extract tile center/velocity into arrays before thread launch
- Replace new ImagePlus(FloatProcessor) debug calls with ShowDoubleFloatArrays.showArrays()
- Fix "centre" -> "center" in comments; naming: nFr for ai.getAndIncrement(), ntile for plain for