Merge pull request #1731 from pierotofy/median

C++ median smoothing filter

SuperBuild/CMakeLists.txt

@@ -251,6 +251,15 @@ externalproject_add(odm_orthophoto
                       ${WIN32_CMAKE_ARGS} ${WIN32_GDAL_ARGS}
 )
 
+externalproject_add(fastrasterfilter
+    GIT_REPOSITORY    https://github.com/OpenDroneMap/FastRasterFilter.git
+    GIT_TAG           main
+    PREFIX            ${SB_BINARY_DIR}/fastrasterfilter
+    SOURCE_DIR        ${SB_SOURCE_DIR}/fastrasterfilter
+    CMAKE_ARGS        -DCMAKE_INSTALL_PREFIX:PATH=${SB_INSTALL_DIR}
+                      ${WIN32_CMAKE_ARGS} ${WIN32_GDAL_ARGS}
+)
+
 externalproject_add(lastools
     GIT_REPOSITORY    https://github.com/OpenDroneMap/LAStools.git
     GIT_TAG           250

opendm/dem/commands.py

@@ -5,8 +5,6 @@ import numpy
 import math
 import time
 import shutil
-import functools
-import threading
 import glob
 import re
 from joblib import delayed, Parallel
@@ -15,11 +13,9 @@ from opendm import point_cloud
 from opendm import io
 from opendm import system
 from opendm.concurrency import get_max_memory, parallel_map, get_total_memory
-from scipy import ndimage
 from datetime import datetime
 from opendm.vendor.gdal_fillnodata import main as gdal_fillnodata
 from opendm import log
-import threading
 
 from .ground_rectification.rectify import run_rectification
 from . import pdal
@@ -237,106 +233,31 @@ def compute_euclidean_map(geotiff_path, output_path, overwrite=False):
         return output_path
 
 
-def median_smoothing(geotiff_path, output_path, smoothing_iterations=1, window_size=512, num_workers=1):
+def median_smoothing(geotiff_path, output_path, window_size=512, num_workers=1, radius=4):
     """ Apply median smoothing """
     start = datetime.now()
 
     if not os.path.exists(geotiff_path):
         raise Exception('File %s does not exist!' % geotiff_path)
 
-    # Prepare temporary files
+    kwargs = {
-    folder_path, output_filename = os.path.split(output_path)
+        'input': geotiff_path,
-    basename, ext = os.path.splitext(output_filename)
+        'output': output_path,
-
+        'window': window_size,
-    output_dirty_in = os.path.join(folder_path, "{}.dirty_1{}".format(basename, ext))
+        'radius': radius,
-    output_dirty_out = os.path.join(folder_path, "{}.dirty_2{}".format(basename, ext))
+    }
-
+    system.run('fastrasterfilter "{input}" '
-    log.ODM_INFO('Starting smoothing...')
+                '--output "{output}" '
-
+                '--window-size {window} '
-    with rasterio.open(geotiff_path, num_threads=num_workers) as img, rasterio.open(output_dirty_in, "w+", BIGTIFF="IF_SAFER", num_threads=num_workers, **img.profile) as imgout, rasterio.open(output_dirty_out, "w+", BIGTIFF="IF_SAFER", num_threads=num_workers, **img.profile) as imgout2:
+                '--radius {radius} '
-        nodata = img.nodatavals[0]
+                '--co TILED=YES '
-        dtype = img.dtypes[0]
+                '--co BIGTIFF=IF_SAFER '
-        shape = img.shape
+                '--co COMPRESS=DEFLATE '.format(**kwargs), env_vars={'OMP_NUM_THREADS': num_workers})
-        for i in range(smoothing_iterations):
-            log.ODM_INFO("Smoothing iteration %s" % str(i + 1))
-            rows, cols = numpy.meshgrid(numpy.arange(0, shape[0], window_size), numpy.arange(0, shape[1], window_size))
-            rows = rows.flatten()
-            cols = cols.flatten()
-            rows_end = numpy.minimum(rows + window_size, shape[0])
-            cols_end= numpy.minimum(cols + window_size, shape[1])
-            windows = numpy.dstack((rows, cols, rows_end, cols_end)).reshape(-1, 4)
-
-            filt = functools.partial(ndimage.median_filter, size=9, output=dtype, mode='nearest')
-
-            # We cannot read/write to the same file from multiple threads without causing race conditions. 
-            # To safely read/write from multiple threads, we use a lock to protect the DatasetReader/Writer.
-            read_lock = threading.Lock()
-            write_lock = threading.Lock()
-
-            # threading backend and GIL released filter are important for memory efficiency and multi-core performance
-            Parallel(n_jobs=num_workers, backend='threading')(delayed(window_filter_2d)(img, imgout, nodata , window, 9, filt, read_lock, write_lock) for window in windows)
-
-            # Between each iteration we swap the input and output temporary files
-            #img_in, img_out = img_out, img_in
-            if (i == 0):
-                img = imgout
-                imgout = imgout2
-            else:
-                img, imgout = imgout, img
-    
-    # If the number of iterations was even, we need to swap temporary files
-    if (smoothing_iterations % 2 != 0):
-        output_dirty_in, output_dirty_out = output_dirty_out, output_dirty_in
-
-    # Cleaning temporary files
-    if os.path.exists(output_dirty_out):
-        os.replace(output_dirty_out, output_path)
-    if os.path.exists(output_dirty_in):
-        os.remove(output_dirty_in)
 
     log.ODM_INFO('Completed smoothing to create %s in %s' % (output_path, datetime.now() - start))
     return output_path
 
 
-def window_filter_2d(img, imgout, nodata, window, kernel_size, filter, read_lock, write_lock):
-    """
-    Apply a filter to dem within a window, expects to work with kernal based filters
-
-    :param img: path to the geotiff to filter
-    :param imgout: path to write the giltered geotiff to. It can be the same as img to do the modification in place.
-    :param window: the window to apply the filter, should be a list contains row start, col_start, row_end, col_end
-    :param kernel_size: the size of the kernel for the filter, works with odd numbers, need to test if it works with even numbers
-    :param filter: the filter function which takes a 2d array as input and filter results as output.
-    :param read_lock: threading lock for the read operation
-    :param write_lock: threading lock for the write operation
-    """
-    shape = img.shape[:2]
-    if window[0] < 0 or window[1] < 0 or window[2] > shape[0] or window[3] > shape[1]:
-        raise Exception('Window is out of bounds')
-    expanded_window = [ max(0, window[0] - kernel_size // 2), max(0, window[1] - kernel_size // 2), min(shape[0], window[2] + kernel_size // 2), min(shape[1], window[3] + kernel_size // 2) ]
-
-    # Read input window
-    width = expanded_window[3] - expanded_window[1]
-    height = expanded_window[2] - expanded_window[0]
-    rasterio_window = rasterio.windows.Window(col_off=expanded_window[1], row_off=expanded_window[0], width=width, height=height)
-    with read_lock:
-        win_arr = img.read(indexes=1, window=rasterio_window)
-
-    # Should have a better way to handle nodata, similar to the way the filter algorithms handle the border (reflection, nearest, interpolation, etc).
-    # For now will follow the old approach to guarantee identical outputs
-    nodata_locs = win_arr == nodata
-    win_arr = filter(win_arr)
-    win_arr[nodata_locs] = nodata
-    win_arr = win_arr[window[0] - expanded_window[0] : window[2] - expanded_window[0], window[1] - expanded_window[1] : window[3] - expanded_window[1]]
-
-    # Write output window
-    width = window[3] - window[1]
-    height = window[2] - window[0]
-    rasterio_window = rasterio.windows.Window(col_off=window[1], row_off=window[0], width=width, height=height)
-    with write_lock:
-        imgout.write(win_arr, indexes=1, window=rasterio_window)
-
-
 def get_dem_radius_steps(stats_file, steps, resolution, multiplier = 1.0):
     radius_steps = [point_cloud.get_spacing(stats_file, resolution) * multiplier]
     for _ in range(steps - 1):