Remove align

plugins/changedetection/align/align_and_rasterize.py

@@ -1,8 +0,0 @@
-from .rasterize import rasterize_cloud
-from .cloud_compare import run_icp
-
-def rasterize(point_cloud_path, dem_type, raster_path):
-    rasterize_cloud(point_cloud_path, dem_type, raster_path)
-
-def align(reference_pcl_path, pcl_to_modify_path, output_path):
-    run_icp(reference_pcl_path, pcl_to_modify_path, output_path)

plugins/changedetection/align/cloud_compare.py

@@ -1,24 +0,0 @@
-import subprocess, shlex, os, glob
-
-def run_icp(reference_pcl_path, pcl_to_modify_path, output_path):
-    command = __icp_command(reference_pcl_path, pcl_to_modify_path, output_path)
-    __execute(command)
-    for txt in glob.glob(os.path.dirname(pcl_to_modify_path) + '/*.txt'):
-        os.remove(txt)
-
-def __execute(command):
-    # subprocess.call(shlex.split(command), stdout=subprocess.DEVNULL, stderr=subprocess.DEVNULL)
-    subprocess.call(shlex.split(command))
-
-def __icp_command(reference_pcl_path, pcl_to_modify_path, output_path):
-    return """CloudCompare \
-            -SILENT \
-            -AUTO_SAVE OFF \
-            -C_EXPORT_FMT LAS \
-            -O {moving} \
-            -O {reference} \
-            -ICP \
-            -SAVE_CLOUDS FILE '{output_path}'""".format(
-                reference=reference_pcl_path,
-                moving=pcl_to_modify_path,
-                output_path=output_path)

plugins/changedetection/align/pdal/pdal.py

@@ -1,29 +0,0 @@
-import os, tempfile
-from string import Template
-import subprocess
-
-def rasterize_dtm(input_pcl, output_tif, resolution = 0.1):
-    __run('dtm_rasterize', fin=input_pcl, output_tif=output_tif, resolution=resolution)
-
-def rasterize_dsm(input_pcl, output_tif, resolution = 0.1):
-    __run('dsm_rasterize', fin=input_pcl, output_tif=output_tif, resolution=resolution)
-
-def info(input_pcl):
-    command = 'pdal info -i \'{}\' --enumerate Classification'.format(input_pcl)
-    output = subprocess.check_output(command, shell=True)
-    return output
-
-def __run(template_name, **kwparams):
-    template = __get_template(template_name)
-    pipeline_definition = template.substitute(kwparams)
-    with tempfile.NamedTemporaryFile(suffix='.json', mode='w') as pipeline:
-        pipeline.write(pipeline_definition)
-        pipeline.flush()
-        command = 'pdal pipeline \'{}\''.format(pipeline.name)
-        output = subprocess.check_output(command, shell=True)
-
-
-def __get_template(template_name):
-    dir = os.path.dirname(os.path.realpath(__file__))
-    with open(os.path.join(dir, 'pipelines', template_name + '.json')) as pipeline_template:
-        return Template(pipeline_template.read())

plugins/changedetection/align/pdal/pipelines/dsm_rasterize.json

@@ -1,14 +0,0 @@
-[
-    "${fin}",
-    {
-        "type": "filters.range",
-        "limits": "Classification![2:2]"
-    },
-    {
-        "type": "writers.gdal",
-        "resolution": ${resolution},
-        "filename": "${output_tif}",
-        "output_type": "max",
-        "data_type": "float"
-    }
-]

plugins/changedetection/align/pdal/pipelines/dtm_rasterize.json

@@ -1,14 +0,0 @@
-[
-    "${fin}",
-    {
-        "type": "filters.range",
-        "limits": "Classification[2:2]"
-    },
-    {
-        "type": "writers.gdal",
-        "resolution": ${resolution},
-        "filename": "${output_tif}",
-        "output_type": "idw",
-        "data_type": "float"
-    }
-]

plugins/changedetection/align/rasterize.py

@@ -1,52 +0,0 @@
-import rasterio as rio
-from scipy import ndimage
-import  tempfile, subprocess, os
-import json
-from .pdal import pdal
-
-def rasterize_cloud(point_cloud_path, dem_type, raster_path):
-    __verify_point_cloud_was_classified(point_cloud_path)
-    with tempfile.NamedTemporaryFile(suffix='.tif') as temp_tif:
-        if dem_type == 'dtm':
-            pdal.rasterize_dtm(point_cloud_path, temp_tif.name)
-        else:
-            pdal.rasterize_dsm(point_cloud_path, temp_tif.name)
-        __fill(temp_tif.name, raster_path)
-
-def __verify_point_cloud_was_classified(point_cloud_path):
-    info = pdal.info(point_cloud_path)
-    info = json.loads(info)
-    dimensions = info['stats']['statistic']
-    classification = [ dimension for dimension in dimensions if dimension['name'] == 'Classification'][0]
-    if 2 not in classification['values']:
-        raise Exception("One of the tasks' point cloud wasn't classified. Please run again with 'pc-classify'.")
-
-def __fill(fin, fout):
-    subprocess.run(['gdal_fillnodata.py', '-q', '-b', '1', '-of', 'GTiff', fin, fin])
-    __median_smoothing(fin, fout)
-
-def __median_smoothing(input_path, output_path, smoothing_iterations = 1):
-    with rio.open(input_path) as input_raster:
-        dtype = input_raster.dtypes[0]
-        arr = input_raster.read(1)
-        mask = arr == input_raster.nodata
-
-        # Median filter (careful, changing the value 5 might require tweaking)
-        # the lines below. There's another numpy function that takes care of
-        # these edge cases, but it's slower.
-        for i in range(smoothing_iterations):
-            arr = ndimage.median_filter(arr, size = 5, output = dtype)
-
-        # Fill corner points with nearest value
-        if arr.shape >= (4, 4):
-            arr[0][:2] = arr[1][0] = arr[1][1]
-            arr[0][-2:] = arr[1][-1] = arr[2][-1]
-            arr[-1][:2] = arr[-2][0] = arr[-2][1]
-            arr[-1][-2:] = arr[-2][-1] = arr[-2][-2]
-
-        # Median filter leaves a bunch of zeros in nodata areas
-        arr[mask == True] = input_raster.nodata
-
-        # write output
-        with rio.open(output_path, 'w', **input_raster.profile) as output_raster:
-            output_raster.write(arr, 1)