cars.pipelines.surface_modeling.surface_modeling

CARS surface modeling pipeline class file

Attributes

PIPELINE

Classes

SurfaceModelingPipeline

SurfaceModelingPipeline

Module Contents

cars.pipelines.surface_modeling.surface_modeling.PIPELINE = 'surface_modeling'

class cars.pipelines.surface_modeling.surface_modeling.SurfaceModelingPipeline(conf, config_dir=None)

Bases: cars.pipelines.pipeline_template.PipelineTemplate

SurfaceModelingPipeline

static _should_run_disparity_to_depth_maps(compute_depth_map, quit_after_dem_generation, quit_after_grid_or_resampling, quit_after_dense_matching)

Return whether disparity_to_depth_maps phase is expected to run.

static _compute_generate_disparity_grids_runs(num_sensor_pairs, has_tie_points_pipeline, has_low_res_dsm, is_first_or_single, use_global_disp_range, quit_after_grid_or_resampling, save_output_dsm, should_run_disparity_to_depth_maps)

Calculate expected runs for generate_disparity_grids.

static _compute_tie_points_runs(has_tie_points_pipeline, quit_after_grid_or_resampling, num_sensor_pairs)

Calculate expected runs for tie_points.

static _compute_disparity_to_depth_maps_runs(should_run_disparity_to_depth_maps, quit_after_triangulation, use_sensor_disp, num_sensor_pairs, save_output_dsm)

Calculate expected runs for disparity_to_depth_maps.

static _compute_rasterize_point_cloud_runs(save_output_dsm, quit_after_rasterization, has_aux_filling)

Calculate expected runs for rasterize_point_cloud.

setup_progress_tracking(parent_pipeline_id=None, tie_points_pipeline_id=None)

Setup progress tracking for surface modeling.

Parameters:

• parent_pipeline_id (int or None) – Optional parent pipeline ID

• tie_points_pipeline_id (int or None) – Optional tie_points pipeline ID pre-created by default pipeline

Returns:

Task ID to pass to orchestrator via set_target_task()

Return type:

int

used_conf

refined_conf

metadata = None

config_dir = None

out_dir

elevation_delta_lower_bound = -500

elevation_delta_upper_bound = 1000

dem_scaling_coeff = None

tie_point_save = False

debug_with_roi

dump_dir

product_format

save_output_dsm

save_output_point_cloud

save_output_dtm

output_level_none

used_classif_values_for_filling = None

phasing

compute_depth_map

save_all_intermediate_data

save_all_point_clouds_by_pair

check_pipeline_conf(conf)

Check pipeline configuration

quit_on_app(app_name)

Returns whether the pipeline should end after the application was called.

Only works if the output_level is empty, so that the control is instead given to

infer_conditions_from_applications(conf)

Fills the condition booleans used later in the pipeline by going through the applications and infering which application we should end the pipeline on.

static check_inputs(conf, config_dir=None)

Check the inputs given

Parameters:

• conf (dict) – configuration of inputs

• config_dir (str) – directory of used json/yaml, if user filled paths with relative paths

Returns:

overloaded inputs

Return type:

dict

save_configurations()

Save used_conf and refined_conf configurations

check_output(inputs, conf, scaling_coeff, bounds)

Check the output given

Parameters:

• conf (dict) – configuration of output

• scaling_coeff (float) – scaling factor for resolution

Returns:

overloader output

Return type:

dict

check_applications(conf)

Check the given configuration for applications, and generates needed applications for pipeline.

Parameters:

conf (dict) – configuration of applications

get_classif_values_filling(inputs)

Get values in classif, used for filling

Parameters:

inputs (dict) – inputs

Returns:

list of values

Return type:

list

check_applications_with_inputs(inputs_conf, application_conf)

Check for each application the input and output configuration consistency

Parameters:

• inputs_conf (dict) – inputs checked configuration

• application_conf (dict) – application checked configuration

sensor_to_disparity()

Creates the disparity map from the sensor images given in the input, by following the CARS pipeline’s steps.

disparity_to_depth_maps()

Creates the depth map from the disparity maps, by following the CARS pipeline’s steps.

rasterize_point_cloud()

Final step of the pipeline: rasterize the point cloud created in the prior steps.

merge_filling_bands(in_filling_path, out_filling_path, aux_filling, dsm_file, invalidity_mask_file, local_orchestrator=None, tile_size=10000)

Merge filling bands to get mono band in output

merge_classif_bands(classif_path, aux_classif, dsm_file)

Merge classif bands to get mono band in output

merge_invalidity_mask_bands(invalidity_mask_path, dsm_file)

Merge invalidity mask bands to get mono band in output

preprocess_depth_maps()

Adds multiple processing steps to the depth maps : Merging. Creates the point cloud that will be rasterized in the last step of the pipeline.

final_cleanup()

Clean temporary files and directory at the end of cars processing

run(args=None, which_resolution='single', working_res=1, res_factor=None, log_dir=None, previous_out_dir=None, parent_pipeline_id=None, tie_points_pipeline_id=None)

Run pipeline

Parameters:

parent_pipeline_id – Optional parent pipeline ID if nested