stable preset

Stabilization quality: full res + more features + CLAHE + stricter matching; slower, more accurate.

Use it with -c stable, or start from a local copy with geotrax config copy. Generated from stable.yaml.

Changed keys

Key	default	stable
stabilo.clahe	false	true
stabilo.downsample_ratio	0.5	1.0
stabilo.max_features	2000	4000
stabilo.filter_ratio	0.9	0.8

Full diff against default.yaml

--- default.yaml
+++ stable.yaml
@@ -1,3 +1,13 @@
-# geo-trax pipeline configuration (default.yaml) — the single, self-contained config for
-# the entire pipeline. There are no linked sub-config files; every setting lives here.
+# geo-trax pipeline configuration — "stable" preset (stable.yaml).
+#
+# A stabilization-quality preset: full-resolution stabilization with more keypoints, CLAHE
+# pre-processing, and a stricter match filter. Produces higher-fidelity feature extraction and
+# more precise bounding-box stabilization, at the cost of slower processing. Use when
+# stabilization quality is the priority (e.g. shaky footage). The ORB detector is retained.
+#
+# Differs from default.yaml:
+#   stabilo.downsample_ratio: 0.5   -> 1.0
+#   stabilo.max_features:     2000  -> 4000
+#   stabilo.clahe:            false -> true
+#   stabilo.filter_ratio:     0.9   -> 0.8
 #
@@ -6,6 +16,6 @@
 #   geotrax config show           # show where this file lives and which presets are available
-#   geotrax config show default   # print this file's full contents to the terminal
+#   geotrax config show stable    # print this file's full contents to the terminal
 #   geotrax config copy           # copy the presets into the current directory as <name>_copy.yaml
 # Then pass your edited copy to any command with -c, e.g.:
-#   geotrax extract video.mp4 -c default_copy.yaml
+#   geotrax extract video.mp4 -c stable_copy.yaml
 #
@@ -121,4 +131,4 @@
   # Image pre-processing
-  clahe: false          # Apply Contrast Limited Adaptive Histogram Equalization before feature detection; [true, false]
-  downsample_ratio: 0.5 # Resize input frames by this factor before processing; lower = faster but fewer features; (0.0, 1.0]
+  clahe: true           # Apply Contrast Limited Adaptive Histogram Equalization before feature detection; [true, false] — enabled (stability preset): more features in low-contrast regions
+  downsample_ratio: 1.0 # Resize input frames by this factor before processing; lower = faster but fewer features; (0.0, 1.0]

@@ -126,3 +136,3 @@
   detector_name: 'orb'  # Feature detector/descriptor; classical: [orb, sift, rsift, brisk, kaze, akaze], learned: [xfeat, disk, dedode, keynet, loftr] — CLI: --stab-detector
-  max_features: 2000    # Maximum keypoints kept per frame, the strongest by response; higher = more robust but slower; [0, inf)
+  max_features: 4000    # Maximum keypoints kept per frame, the strongest by response; higher = more robust but slower; [0, inf) — raised (stability preset) for a more robust homography
   ref_multiplier: 2.0   # Multiplier applied to max_features for the reference frame only; (0.0, inf)
@@ -134,3 +144,3 @@
   filter_type: 'ratio'  # Post-match filter; choices: [none, ratio (Lowe), distance]
-  filter_ratio: 0.9     # Lowe ratio-test threshold; lower = stricter (fewer but better matches); (0.0, 1.0]
+  filter_ratio: 0.8     # Lowe ratio-test threshold; lower = stricter (fewer but better matches); (0.0, 1.0] — stricter (stability preset): keep only the best matches