generated from ashleve/lightning-hydra-template
-
Notifications
You must be signed in to change notification settings - Fork 3
Commit
This commit does not belong to any branch on this repository, and may belong to a fork outside of the repository.
- Loading branch information
1 parent
ad12986
commit b816349
Showing
15 changed files
with
286 additions
and
124 deletions.
There are no files selected for viewing
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| Original file line number | Diff line number | Diff line change |
|---|---|---|
| @@ -0,0 +1,15 @@ | ||
| _target_: lidar_prod.tasks.building_completion.BuildingCompletor | ||
|
|
||
| data_format: ${data_format} | ||
|
|
||
| # TODO: uncomment when min_building_proba_relaxation_if_bd_uni_overlay is specified <1 to be consistent everywhere. | ||
| # min_building_proba: ${building_validation.application.rules.min_confidence_confirmation} | ||
|
|
||
| min_building_proba: 0.5 | ||
| min_building_proba_relaxation_if_bd_uni_overlay: 1.0 # No effect if = 1.0 | ||
|
|
||
| cluster: | ||
| min_points: 10 # including isolated points (in BuildingValidator) and confirmed candidates points. | ||
| tolerance: 1 # meters | ||
| is3d: false # group in 2d for better detection | ||
|
|
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| Original file line number | Diff line number | Diff line change |
|---|---|---|
| @@ -1,14 +1,12 @@ | ||
| _target_: lidar_prod.tasks.building_identification.BuildingIdentifier | ||
|
|
||
| candidate_buildings_codes: ${data_format.codes.candidates.building} | ||
|
|
||
| data_format: ${data_format} | ||
|
|
||
| min_building_proba: ${building_validation.application.rules.min_confidence_confirmation} | ||
| min_building_proba_multiplier_if_bd_uni_overlay: 1.0 | ||
| min_building_proba_relaxation_if_bd_uni_overlay: 1.0 | ||
|
|
||
| cluster: | ||
| min_points: 200 | ||
| min_points: 200 # Large so that small artefact are ignored | ||
| tolerance: 1 # meters | ||
| is3d: false # group in 2d for better detection | ||
|
|
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| Original file line number | Diff line number | Diff line change |
|---|---|---|
| @@ -0,0 +1,148 @@ | ||
| import logging | ||
| import os | ||
| import os.path as osp | ||
| from tempfile import TemporaryDirectory | ||
| import pdal | ||
| import laspy | ||
| from tqdm import tqdm | ||
|
|
||
| from lidar_prod.tasks.utils import split_idx_by_dim | ||
|
|
||
| log = logging.getLogger(__name__) | ||
|
|
||
|
|
||
| class BuildingCompletor: | ||
| """Logic of building completion. | ||
| Some points were too isolated for BuildingValidator to consider them. | ||
| We will update their classification based on their probability as well as their surrounding: | ||
| - We select points that have p>=0.5 | ||
| - We perform vertical (XY) clustering including these points as well as confirmed buildings. | ||
| - In the resulting groups, if there are some confirmed buildings, previously isolated points are | ||
| considered to be parts of the same building and their class is updated accordingly. | ||
| """ | ||
|
|
||
| def __init__( | ||
| self, | ||
| min_building_proba: float = 0.75, | ||
| min_building_proba_relaxation_if_bd_uni_overlay: float = 1.0, | ||
| cluster=None, | ||
| data_format=None, | ||
| ): | ||
| self.cluster = cluster | ||
| self.min_building_proba = min_building_proba | ||
| self.min_building_proba_relaxation_if_bd_uni_overlay = ( | ||
| min_building_proba_relaxation_if_bd_uni_overlay | ||
| ) | ||
| self.data_format = data_format | ||
| self.codes = data_format.codes.building # easier access | ||
|
|
||
| def run(self, in_f: str, out_f: str): | ||
| """Application. | ||
| Transform cloud at `in_f` following building completion logic, and save it to | ||
| `out_f` | ||
| Args: | ||
| in_f (str): path to input LAS file, output of BuildingValidator | ||
| out_f (str): path for saving updated LAS file. | ||
| Returns: | ||
| _type_: returns `out_f` for potential terminal piping. | ||
| """ | ||
| log.info(f"Applying Building Completion to file \n{in_f}") | ||
| log.info( | ||
| "Completion of building with relatively distant points that have high enough probability" | ||
| ) | ||
| with TemporaryDirectory() as td: | ||
| temp_f = osp.join(td, osp.basename(in_f)) | ||
| self.prepare(in_f, temp_f) | ||
| self.update(temp_f, out_f) | ||
| return out_f | ||
|
|
||
| def prepare(self, in_f: str, out_f: str): | ||
| f"""Prepare for building completion. | ||
| Identify candidates that were not clustered together by the BuildingValidator, but that | ||
| have high enough probability. Then, cluster them together with previously confirmed buildings. | ||
| Cluster parameters are relaxed (2D, with high tolerance). | ||
| If a cluster contains some confirmed points, the others are considered to belong to the same building | ||
| and they will be confirmed as well. | ||
| Args: | ||
| in_f (str): input LAS | ||
| out_f (str): output, prepared LAS with a new `{self.data_format.las_dimensions.ClusterID_isolated_plus_confirmed}` | ||
| dimension. | ||
| """ | ||
| pipeline = pdal.Pipeline() | ||
| pipeline |= pdal.Reader( | ||
| in_f, | ||
| type="readers.las", | ||
| # nosrs=True, | ||
| # override_srs=self.data_format.crs_prefix + str(self.data_format.crs), | ||
| ) | ||
| candidates = ( | ||
| f"({self.data_format.las_dimensions.candidate_buildings_flag} == 1)" | ||
| ) | ||
|
|
||
| where_not_clustered = ( | ||
| f"{self.data_format.las_dimensions.ClusterID_candidate_building} == 0" | ||
| ) | ||
|
|
||
| # P above threshold | ||
| p_heq_threshold = f"(building>={self.min_building_proba})" | ||
|
|
||
| # P above relaxed threshold when under BDUni | ||
| under_bd_uni = f"({self.data_format.las_dimensions.uni_db_overlay} > 0)" | ||
| p_heq_relaxed_threshold = f"(building>={self.min_building_proba * self.min_building_proba_relaxation_if_bd_uni_overlay})" | ||
| p_heq_threshold_under_bd_uni = f"({p_heq_relaxed_threshold} && {under_bd_uni})" | ||
|
|
||
| # Candidates that where clustered by BuildingValidator but have high enough probability. | ||
| not_clustered_but_with_high_p = f"{candidates} && {where_not_clustered} && ({p_heq_threshold} || {p_heq_threshold_under_bd_uni})" | ||
| confirmed_buildings = ( | ||
| f"Classification == {self.data_format.codes.building.final.building}" | ||
| ) | ||
|
|
||
| where = f"{not_clustered_but_with_high_p} || {confirmed_buildings}" | ||
| pipeline |= pdal.Filter.cluster( | ||
| min_points=self.cluster.min_points, | ||
| tolerance=self.cluster.tolerance, | ||
| is3d=self.cluster.is3d, | ||
| where=where, | ||
| ) | ||
| # Always move and reset ClusterID to avoid conflict with later tasks. | ||
| pipeline |= pdal.Filter.ferry( | ||
| dimensions=f"{self.data_format.las_dimensions.cluster_id}=>{self.data_format.las_dimensions.ClusterID_isolated_plus_confirmed}" | ||
| ) | ||
| pipeline |= pdal.Filter.assign( | ||
| value=f"{self.data_format.las_dimensions.cluster_id} = 0" | ||
| ) | ||
| pipeline |= pdal.Writer( | ||
| type="writers.las", filename=out_f, forward="all", extra_dims="all" | ||
| ) | ||
| os.makedirs(osp.dirname(out_f), exist_ok=True) | ||
| pipeline.execute() | ||
|
|
||
| def update(self, prepared_f: str, out_f: str): | ||
| """ | ||
| Args: | ||
| in_f (str): input, prepared LAS | ||
| out_f (str): output LAS, with updated Classification dimension. | ||
| """ | ||
| las = laspy.read(prepared_f) | ||
| _clf = self.data_format.las_dimensions.classification | ||
| _cid = self.data_format.las_dimensions.ClusterID_isolated_plus_confirmed | ||
| # 2) Decide at the group-level | ||
| split_idx = split_idx_by_dim(las[_cid]) | ||
| # Isolated/confirmed groups have a cluster index > 0 | ||
| split_idx = split_idx[1:] | ||
| for pts_idx in tqdm( | ||
| split_idx, desc="Complete buildings with isolated points", unit="grp" | ||
| ): | ||
| pts = las.points[pts_idx] | ||
| if self.codes.final.building in pts[_clf]: | ||
| las[_clf][pts_idx] = self.codes.final.building | ||
| os.makedirs(osp.dirname(out_f), exist_ok=True) | ||
| las.write(out_f) |
Oops, something went wrong.