code_public/lidar_rendu
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

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

Browse Source 
- 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>
This commit is contained in:
Jacquin Antoine
2026-05-14 00:50:45 +02:00
parent eac482874d
commit 1cf8e1752f
6 changed files with 66 additions and 63 deletions
Download Patch File Download Diff File Expand all files Collapse all files

30
lidar_pipeline/visualizations.py
View File

@ -100,8 +100,18 @@ def _filter_nanaware_from_filled(shared, filter_func, *args, **kwargs):
return result
def _save_tif(output_path, data, transform, crs, dtype='float32', count=1, nodata=None):
"""Helper to save a 2D or 3D array as GeoTIFF."""
def _save_tif(output_path, data, transform, crs, dtype='float32', count=1, nodata=None, nan_mask=None):
"""Helper to save a 2D or 3D array as GeoTIFF.
Args:
nan_mask: Optional boolean mask (True=NaN) to apply before saving.
Restores NaN zones in gradient-derived products that were
computed on the filled DEM.
"""
if nan_mask is not None:
data = np.array(data, dtype=dtype, copy=True)
data[nan_mask] = np.nan
# Auto-detect nodata for float types with NaN
if nodata is None and dtype.startswith('float') and np.any(np.isnan(data)):
nodata = float('nan')
@ -238,7 +248,9 @@ def generate_hillshade(dem_file, basename, vis_dir, resolution, shared=None):
combined_hillshade = xp.mean(xp.array(hillshades), axis=0)
slope_shaded = cos_slope
combined = 0.7 * combined_hillshade + 0.3 * slope_shaded
_save_tif(output, to_cpu(combined), transform, crs)
nan_mask = shared.nan_mask if shared else np.isnan(to_cpu(dem_np))
_save_tif(output, to_cpu(combined), transform, crs, nan_mask=nan_mask)
logger.info(f" ✓ Hillshade terminé ({time.time()-t0:.1f}s){gpu_tag}")
return output
except Exception as e:
@ -258,6 +270,7 @@ def generate_slope(dem_file, basename, vis_dir, resolution, shared=None):
transform = shared.transform
crs = shared.crs
slope = shared.slope_deg
nan_mask = shared.nan_mask
if HAS_GPU:
slope = to_gpu(slope)
else:
@ -265,7 +278,8 @@ def generate_slope(dem_file, basename, vis_dir, resolution, shared=None):
dem = to_gpu(dem_np)
dy, dx = xp.gradient(dem)
slope = xp.arctan(xp.sqrt(dx**2 + dy**2)) * 180 / xp.pi
_save_tif(output, to_cpu(slope) if HAS_GPU else slope, transform, crs)
nan_mask = np.isnan(dem_np)
_save_tif(output, to_cpu(slope) if HAS_GPU else slope, transform, crs, nan_mask=nan_mask)
logger.info(f" ✓ Pente terminée ({time.time()-t0:.1f}s){gpu_tag}")
return output
except Exception as e:
@ -285,6 +299,7 @@ def generate_aspect(dem_file, basename, vis_dir, resolution, shared=None):
transform = shared.transform
crs = shared.crs
aspect = shared.aspect
nan_mask = shared.nan_mask
if HAS_GPU:
aspect = to_gpu(aspect)
else:
@ -293,7 +308,8 @@ def generate_aspect(dem_file, basename, vis_dir, resolution, shared=None):
dy, dx = xp.gradient(dem)
aspect = xp.arctan2(dy, dx) * 180 / xp.pi
aspect = xp.mod(aspect, 360)
_save_tif(output, to_cpu(aspect) if HAS_GPU else aspect, transform, crs)
nan_mask = np.isnan(dem_np)
_save_tif(output, to_cpu(aspect) if HAS_GPU else aspect, transform, crs, nan_mask=nan_mask)
logger.info(f" ✓ Aspect terminé ({time.time()-t0:.1f}s){gpu_tag}")
return output
except Exception as e:
@ -314,6 +330,7 @@ def generate_curvature(dem_file, basename, vis_dir, resolution, shared=None):
crs = shared.crs
dx = shared.dx
dy = shared.dy
nan_mask = shared.nan_mask
if HAS_GPU:
dx = to_gpu(dx)
dy = to_gpu(dy)
@ -321,10 +338,11 @@ def generate_curvature(dem_file, basename, vis_dir, resolution, shared=None):
dem_np, transform, crs = _read_dem(dem_file)
dem = to_gpu(dem_np)
dy, dx = xp.gradient(dem)
nan_mask = np.isnan(dem_np)
d2z_dx2 = xp.gradient(dx, axis=1)
d2z_dy2 = xp.gradient(dy, axis=0)
curvature = (d2z_dx2 + d2z_dy2) / 2
_save_tif(output, to_cpu(curvature), transform, crs)
_save_tif(output, to_cpu(curvature), transform, crs, nan_mask=nan_mask)
logger.info(f" ✓ Courbure terminée ({time.time()-t0:.1f}s){gpu_tag}")
return output
except Exception as e: