import os import json import numpy as np import math from repoze.lru import lru_cache from opendm import log def rounded_gsd(reconstruction_json, default_value=None, ndigits=0, ignore_gsd=False): """ :param reconstruction_json path to OpenSfM's reconstruction.json :return GSD value rounded. If GSD cannot be computed, or ignore_gsd is set, it returns a default value. """ if ignore_gsd: return default_value gsd = opensfm_reconstruction_average_gsd(reconstruction_json) if gsd is not None: return round(gsd, ndigits) else: return default_value def image_max_size(photos, target_resolution, reconstruction_json, gsd_error_estimate = 0.5, ignore_gsd=False): """ :param photos images database :param target_resolution resolution the user wants have in cm / pixel :param reconstruction_json path to OpenSfM's reconstruction.json :param gsd_error_estimate percentage of estimated error in the GSD calculation to set an upper bound on resolution. :param ignore_gsd if set to True, simply return the largest side of the largest image in the images database. :return A dimension in pixels calculated by taking the image_scale_factor and applying it to the size of the largest image. Returned value is never higher than the size of the largest side of the largest image. """ max_width = 0 max_height = 0 if ignore_gsd: isf = 1.0 else: isf = image_scale_factor(target_resolution, reconstruction_json, gsd_error_estimate) for p in photos: max_width = max(p.width, max_width) max_height = max(p.height, max_height) return int(math.ceil(max(max_width, max_height) * isf)) def image_scale_factor(target_resolution, reconstruction_json, gsd_error_estimate = 0.5): """ :param target_resolution resolution the user wants have in cm / pixel :param reconstruction_json path to OpenSfM's reconstruction.json :param gsd_error_estimate percentage of estimated error in the GSD calculation to set an upper bound on resolution. :return A down-scale (<= 1) value to apply to images to achieve the target resolution by comparing the current GSD of the reconstruction. If a GSD cannot be computed, it just returns 1. Returned scale values are never higher than 1. """ gsd = opensfm_reconstruction_average_gsd(reconstruction_json) if gsd is not None and target_resolution > 0: gsd = gsd * (1 + gsd_error_estimate) return min(1, gsd / target_resolution) else: return 1 def cap_resolution(resolution, reconstruction_json, gsd_error_estimate = 0.1, ignore_gsd=False): """ :param resolution resolution in cm / pixel :param reconstruction_json path to OpenSfM's reconstruction.json :param gsd_error_estimate percentage of estimated error in the GSD calculation to set an upper bound on resolution. :param ignore_gsd when set to True, forces the function to just return resolution. :return The max value between resolution and the GSD computed from the reconstruction. If a GSD cannot be computed, or ignore_gsd is set to True, it just returns resolution. Units are in cm / pixel. """ if ignore_gsd: return resolution gsd = opensfm_reconstruction_average_gsd(reconstruction_json) if gsd is not None: gsd = gsd * (1 - gsd_error_estimate) if gsd > resolution: log.ODM_WARNING('Maximum resolution set to GSD - {}% ({} cm / pixel, requested resolution was {} cm / pixel)'.format(gsd_error_estimate * 100, round(gsd, 2), round(resolution, 2))) return gsd else: return resolution else: log.ODM_WARNING('Cannot calculate GSD, using requested resolution of {}'.format(round(resolution, 2))) return resolution @lru_cache(maxsize=None) def opensfm_reconstruction_average_gsd(reconstruction_json): """ Computes the average Ground Sampling Distance of an OpenSfM reconstruction. :param reconstruction_json path to OpenSfM's reconstruction.json :return Ground Sampling Distance value (cm / pixel) or None if a GSD estimate cannot be compute """ if not os.path.isfile(reconstruction_json): raise IOError(reconstruction_json + " does not exist.") with open(reconstruction_json) as f: data = json.load(f) # Calculate median height from sparse reconstruction reconstruction = data[0] point_heights = [] for pointId in reconstruction['points']: point = reconstruction['points'][pointId] point_heights.append(point['coordinates'][2]) ground_height = np.median(point_heights) gsds = [] for shotImage in reconstruction['shots']: shot = reconstruction['shots'][shotImage] if shot['gps_dop'] < 999999: camera = reconstruction['cameras'][shot['camera']] shot_height = shot['translation'][2] focal_ratio = camera.get('focal', camera.get('focal_x')) if not focal_ratio: log.ODM_WARNING("Cannot parse focal values from %s. This is likely an unsupported camera model." % reconstruction_json) return None gsds.append(calculate_gsd_from_focal_ratio(focal_ratio, shot_height - ground_height, camera['width'])) if len(gsds) > 0: mean = np.mean(gsds) if mean > 0: return mean return None def calculate_gsd(sensor_width, flight_height, focal_length, image_width): """ :param sensor_width in millimeters :param flight_height in meters :param focal_length in millimeters :param image_width in pixels :return Ground Sampling Distance >>> round(calculate_gsd(13.2, 100, 8.8, 5472), 2) 2.74 >>> calculate_gsd(13.2, 100, 0, 2000) >>> calculate_gsd(13.2, 100, 8.8, 0) """ if sensor_width != 0: return calculate_gsd_from_focal_ratio(focal_length / sensor_width, flight_height, image_width) else: return None def calculate_gsd_from_focal_ratio(focal_ratio, flight_height, image_width): """ :param focal_ratio focal length (mm) / sensor_width (mm) :param flight_height in meters :param image_width in pixels :return Ground Sampling Distance """ if focal_ratio == 0 or image_width == 0: return None return ((flight_height * 100) / image_width) / focal_ratio