opr.pipelines package

Module for pipelines.

• opr.pipelines.localization package
  • opr.pipelines.localization.aruco
    • ArucoLocalizationPipeline
      • ArucoLocalizationPipeline.aruco_part()
      • ArucoLocalizationPipeline.infer()
      • ArucoLocalizationPipeline.loc_part()
    • pose_to_matrix()
  • opr.pipelines.localization.base
    • LocalizationPipeline
      • LocalizationPipeline.compute_reg_features()
      • LocalizationPipeline.infer()
• opr.pipelines.place_recognition package
  • opr.pipelines.place_recognition.base
    • PlaceRecognitionPipeline
      • PlaceRecognitionPipeline.infer()
  • opr.pipelines.place_recognition.text_labels
    • TextLabelsPlaceRecognitionOCRPipeline
      • TextLabelsPlaceRecognitionOCRPipeline.find_most_similar_id()
      • TextLabelsPlaceRecognitionOCRPipeline.get_labels_by_id()
      • TextLabelsPlaceRecognitionOCRPipeline.infer()
      • TextLabelsPlaceRecognitionOCRPipeline.init_ocr_model()
      • TextLabelsPlaceRecognitionOCRPipeline.normalize_labels()
      • TextLabelsPlaceRecognitionOCRPipeline.remove_stopwords()
    • TextLabelsPlaceRecognitionPipeline
      • TextLabelsPlaceRecognitionPipeline.find_most_similar_id()
      • TextLabelsPlaceRecognitionPipeline.get_labels_by_id()
      • TextLabelsPlaceRecognitionPipeline.infer()
      • TextLabelsPlaceRecognitionPipeline.normalize_labels()
      • TextLabelsPlaceRecognitionPipeline.remove_stopwords()
• opr.pipelines.registration package
  • opr.pipelines.registration.pointcloud
    • PointcloudRegistrationPipeline
      • PointcloudRegistrationPipeline.infer()
    • RansacGlobalRegistrationPipeline
      • RansacGlobalRegistrationPipeline.infer()
    • SequencePointcloudRegistrationPipeline
      • SequencePointcloudRegistrationPipeline.infer()

opr.pipelines.depth_estimation

class opr.pipelines.depth_estimation.DepthEstimationPipeline(model: Module, model_type: str = 'AdelaiDepth', align_type: str = 'average', mode: str = 'indoor', model_weights_path: str | PathLike | None = None, device: str | int | device = 'cuda')

Bases: object

get_depth_with_lidar(image: ndarray, point_cloud: ndarray) → ndarray

Obtain depth estimation from the provided image and point cloud data. :param image: np.ndarray - monocular image :param point_cloud: np.ndarray - sparse lidar point cloud

Returns:

np.ndarray - reconstructed depth map with the same height and width as the input image zs: np.ndarray - z values of the lidar point cloud projected on the iamge errors: np.ndarray - absolute errors of depth reconstruction for the points of the projected lidar point cloud rel_errors: np.ndarray - relative errors of depth reconstruction for the points of the projected lidar point cloud

Return type:

depth

set_camera_matrix(camera_matrix: Dict[str, float])

Set the camera intrinsic matrix for calculations.

set_lidar_to_camera_transform(transform)

Set lidar to camera transform