OpenDroneMap-ODM/stages/odm_orthophoto.py

import os

from opendm import io
from opendm import log
from opendm import system
from opendm import context
from opendm import types
from opendm import gsd
from opendm import orthophoto
from opendm.concurrency import get_max_memory
from opendm.cutline import compute_cutline
from opendm.utils import double_quote
from opendm import pseudogeo
from opendm.multispectral import get_primary_band_name


class ODMOrthoPhotoStage(types.ODM_Stage):
    def process(self, args, outputs):
        tree = outputs['tree']
        reconstruction = outputs['reconstruction']

        # define paths and create working directories
        system.mkdir_p(tree.odm_orthophoto)

        if args.skip_orthophoto:
            log.ODM_WARNING("--skip-orthophoto is set, no orthophoto will be generated")
            return

        if not io.file_exists(tree.odm_orthophoto_tif) or self.rerun():

            resolution = 1.0 / (gsd.cap_resolution(args.orthophoto_resolution, tree.opensfm_reconstruction,
                                                   ignore_gsd=args.ignore_gsd,
                                                   ignore_resolution=(not reconstruction.is_georeferenced()) and args.ignore_gsd,
                                                   has_gcp=reconstruction.has_gcp()) / 100.0)

            # odm_orthophoto definitions
            kwargs = {
                'odm_ortho_bin': context.odm_orthophoto_path,
                'log': tree.odm_orthophoto_log,
                'ortho': tree.odm_orthophoto_render,
                'corners': tree.odm_orthophoto_corners,
                'res': resolution,
                'bands': '',
                'depth_idx': ''
            }

            models = []

            if args.use_3dmesh:
                base_dir = tree.odm_texturing
            else:
                base_dir = tree.odm_25dtexturing
                
            model_file = tree.odm_textured_model_obj

            if reconstruction.multi_camera:
                for band in reconstruction.multi_camera:
                    primary = band['name'] == get_primary_band_name(reconstruction.multi_camera, args.primary_band)
                    subdir = ""
                    if not primary:
                        subdir = band['name'].lower()
                    models.append(os.path.join(base_dir, subdir, model_file))
                kwargs['bands'] = '-bands %s' % (','.join([double_quote(b['name']) for b in reconstruction.multi_camera]))

                # If a RGB band is present, 
                # use bit depth of the first non-RGB band
                depth_idx = None
                all_bands = [b['name'].lower() for b in reconstruction.multi_camera]
                for b in ['rgb', 'redgreenblue']:
                    if b in all_bands:
                        for idx in range(len(all_bands)):
                            if all_bands[idx] != b:
                                depth_idx = idx
                                break
                        break
                
                if depth_idx is not None:
                    kwargs['depth_idx'] = '-outputDepthIdx %s' % depth_idx
            else:
                models.append(os.path.join(base_dir, model_file))

            kwargs['models'] = ','.join(map(double_quote, models))

            # run odm_orthophoto
            system.run('"{odm_ortho_bin}" -inputFiles {models} '
                       '-logFile "{log}" -outputFile "{ortho}" -resolution {res} -verbose '
                       '-outputCornerFile "{corners}" {bands} {depth_idx}'.format(**kwargs))

            # Create georeferenced GeoTiff
            geotiffcreated = False

            if reconstruction.is_georeferenced():
                ulx = uly = lrx = lry = 0.0
                with open(tree.odm_orthophoto_corners) as f:
                    for lineNumber, line in enumerate(f):
                        if lineNumber == 0:
                            tokens = line.split(' ')
                            if len(tokens) == 4:
                                ulx = float(tokens[0]) + \
                                    float(reconstruction.georef.utm_east_offset)
                                lry = float(tokens[1]) + \
                                    float(reconstruction.georef.utm_north_offset)
                                lrx = float(tokens[2]) + \
                                    float(reconstruction.georef.utm_east_offset)
                                uly = float(tokens[3]) + \
                                    float(reconstruction.georef.utm_north_offset)
                log.ODM_INFO('Creating GeoTIFF')

                orthophoto_vars = orthophoto.get_orthophoto_vars(args)

                kwargs = {
                    'ulx': ulx,
                    'uly': uly,
                    'lrx': lrx,
                    'lry': lry,
                    'vars': ' '.join(['-co %s=%s' % (k, orthophoto_vars[k]) for k in orthophoto_vars]),
                    'proj': reconstruction.georef.proj4(),
                    'input': tree.odm_orthophoto_render,
                    'output': tree.odm_orthophoto_tif,
                    'log': tree.odm_orthophoto_tif_log,
                    'max_memory': get_max_memory(),
                }

                system.run('gdal_translate -a_ullr {ulx} {uly} {lrx} {lry} '
                           '{vars} '
                           '-a_srs \"{proj}\" '
                           '--config GDAL_CACHEMAX {max_memory}% '
                           '--config GDAL_TIFF_INTERNAL_MASK YES '
                           '"{input}" "{output}" > "{log}"'.format(**kwargs))

                bounds_file_path = os.path.join(tree.odm_georeferencing, 'odm_georeferenced_model.bounds.gpkg')
                    
                # Cutline computation, before cropping
                # We want to use the full orthophoto, not the cropped one.
                if args.orthophoto_cutline:
                    cutline_file = os.path.join(tree.odm_orthophoto, "cutline.gpkg")

                    compute_cutline(tree.odm_orthophoto_tif, 
                                    bounds_file_path,
                                    cutline_file,
                                    args.max_concurrency,
                                    scale=0.25)

                    orthophoto.compute_mask_raster(tree.odm_orthophoto_tif, cutline_file, 
                                           os.path.join(tree.odm_orthophoto, "odm_orthophoto_cut.tif"),
                                           blend_distance=20, only_max_coords_feature=True)

                orthophoto.post_orthophoto_steps(args, bounds_file_path, tree.odm_orthophoto_tif, tree.orthophoto_tiles)

                # Generate feathered orthophoto also
                if args.orthophoto_cutline:
                    orthophoto.feather_raster(tree.odm_orthophoto_tif, 
                            os.path.join(tree.odm_orthophoto, "odm_orthophoto_feathered.tif"),
                            blend_distance=20
                        )

                geotiffcreated = True
            if not geotiffcreated:
                if io.file_exists(tree.odm_orthophoto_render):
                    pseudogeo.add_pseudo_georeferencing(tree.odm_orthophoto_render)
                    log.ODM_INFO("Renaming %s --> %s" % (tree.odm_orthophoto_render, tree.odm_orthophoto_tif))
                    os.replace(tree.odm_orthophoto_render, tree.odm_orthophoto_tif)
                else:
                    log.ODM_WARNING("Could not generate an orthophoto (it did not render)")
        else:
            log.ODM_WARNING('Found a valid orthophoto in: %s' % tree.odm_orthophoto_tif)

        if args.optimize_disk_space and io.file_exists(tree.odm_orthophoto_render):
            os.remove(tree.odm_orthophoto_render)