Remove PMF, fix NaN in gradient visualizations, fix pos_open/neg_open shared param

- Remove PMF from ground classification options (PDAL recommends SMRF over PMF)
- Auto-detection now uses CSF for urban/complex terrain instead of PMF
- Add z_std > 30m heuristic to auto-select CSF for complex terrain
- Fix pos_open/neg_open lambda missing 'shared' parameter (NameError in workers)
- Fix NaN mask not restored in hillshade, slope, aspect, curvature
  (gradient-based products computed on filled DEM lost NaN transparency)
- Add nan_mask parameter to _save_tif for centralized NaN restoration
- DTM TIF kept by default (no longer deleted after WebP conversion)

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

lidar_pipeline/cli.py

@@ -118,15 +118,15 @@ Exemples:
         help="Reclassifier le sol même si le fichier .las existe déjà"
     )
     parser.add_argument(
-        "--no-keep-tif",
+        "--keep-tif",
         action="store_true",
-        help="Supprimer les fichiers TIFF intermédiaires après conversion WebP (par défaut: conservés)"
+        help="Conserver les fichiers TIFF (DTM + visualisations) pour pouvoir régénérer les WebP sans recalculer"
     )
     parser.add_argument(
         "--ground-classification",
-        choices=["auto", "smrf", "pmf", "csf"],
+        choices=["auto", "smrf", "csf"],
         default="auto",
-        help="Méthode de classification du sol : auto (détection), smrf, pmf, csf (défaut: auto)"
+        help="Méthode de classification du sol : auto (détection), smrf, csf (défaut: auto)"
     )
     parser.add_argument(
         "--file",
@@ -176,7 +176,7 @@ Exemples:
             force=args.force,
             ground_method=args.ground_classification,
             force_classify=args.force_classification,
-            keep_tif=not args.no_keep_tif,
+            keep_tif=args.keep_tif,
         )
 
         # If --file is specified, process only matching files