bboxes_2d_detector
Welcome to the documentation about the detectors for Bounding Boxes 2D.
All functions are defined in details further down the page.
background_subtractor
Copyright (c) 2021-2022 UCLouvain, ICTEAM Licensed under GPL-3.0 [see LICENSE for details] Written by Jonathan Samelson (2021-2022)
- class BackgroundSubtractor(proc_parameters: dict)[source]
Bases:
pytb.detection.bboxes.bboxes_2d_detector.bboxes_2d_detector.BBoxes2DDetector
- __init__(proc_parameters: dict)[source]
Initializes the detector with the given parameters.
- Parameters
proc_parameters (dict) – A dictionary containing the BackgroundSubstractor detector parameters
- detect(frame: numpy.array) pytb.output.bboxes_2d.BBoxes2D [source]
Performs an inference using a background subtraction method on the given frame.
- Parameters
frame (np.array) – The frame to infer detections from a background substractor.
- Returns
A set of 2D bounding boxes identifying the detected objects.
- Return type
detectron2
Copyright (c) 2021-2022 UCLouvain, ICTEAM Licensed under GPL-3.0 [see LICENSE for details] Written by Jonathan Samelson (2021-2022)
- class Detectron2(proc_parameters: dict)[source]
Bases:
pytb.detection.bboxes.bboxes_2d_detector.bboxes_2d_detector.BBoxes2DDetector
- __init__(proc_parameters: dict)[source]
Initializes the detector with the given parameters.
- Parameters
proc_parameters (dict) – A dictionary containing the Detectron2 detector parameters
- detect(org_frame: numpy.array) pytb.output.bboxes_2d.BBoxes2D [source]
Performs a Detectron2 inference on the given frame.
- Parameters
frame (np.array) – The frame to infer Detectron2 detections
- Returns
A set of 2D bounding boxes identifying the detected objects.
- Return type
faster_rcnn
Copyright (c) 2021-2022 UCLouvain, ICTEAM Licensed under GPL-3.0 [see LICENSE for details] Written by Jonathan Samelson and Benoît Gérin (2022)
- class FasterRCNN(proc_parameters: dict)[source]
Bases:
pytb.detection.bboxes.bboxes_2d_detector.bboxes_2d_detector.BBoxes2DDetector
- __init__(proc_parameters: dict)[source]
Initializes the detector with the given parameters.
- Parameters
proc_parameters (dict) – A dictionary containing the FasterRCNN parameters.
- detect(frame: numpy.array) pytb.output.bboxes_2d.BBoxes2D [source]
Performs a Faster-RCNN inference on the given frame.
- Parameters
frame (np.array) – The frame to infer Faster-RCNN detections
- Returns
A set of 2D bounding boxes identifying the detected objects.
- Return type
- _detect_torch_resnet50_pretrained(org_frame) pytb.output.bboxes_2d.BBoxes2D [source]
Performs the inference using the implementation PyTorch Resnet50.
- Parameters
org_frame (np.array) – The frame to infer Faster-RCNN detections.
mask_rcnn
Copyright (c) 2021-2022 UCLouvain, ICTEAM Licensed under GPL-3.0 [see LICENSE for details] Written by Jonathan Samelson and Benoît Gérin (2022)
- class MaskRCNN(proc_parameters: dict)[source]
Bases:
pytb.detection.bboxes.bboxes_2d_detector.bboxes_2d_detector.BBoxes2DDetector
- __init__(proc_parameters: dict)[source]
Initializes the detector with the given parameters.
- Parameters
proc_parameters (dict) – A dictionary containing the MaskRCNN parameters.
- detect(frame: numpy.ndarray) pytb.output.bboxes_2d.BBoxes2D [source]
Performs a Mask-RCNN inference on the given frame.
- Parameters
frame (np.ndarray) – The frame to infer Mask-RCNN detections
- Returns
A set of 2D bounding boxes identifying the detected objects.
- Return type
- _detect_torch_resnet50_pretrained(org_frame) pytb.output.bboxes_2d.BBoxes2D [source]
Performs the inference using the implementation PyTorch Resnet50.
- Parameters
detections. (The frame to infer Mask-RCNN) –
yolo_2_3_4
- class YOLO4(proc_parameters: dict)[source]
Bases:
pytb.detection.bboxes.bboxes_2d_detector.bboxes_2d_detector.BBoxes2DDetector
- __init__(proc_parameters: dict)[source]
This class can be used for YOLO v2, v3, v4 models from Darknet.
Initializes the detector with the given parameters.
- Parameters
proc_parameters (dict) – A dictionary containing the YOLO detector parameters
- detect(frame: numpy.array) pytb.output.bboxes_2d.BBoxes2D [source]
Performs a YOLO inference on the given frame.
- Parameters
frame (np.array) – The frame to infer YOLO detections.
- Returns
A set of 2D bounding boxes identifying the detected objects.
- Return type
- _detect_cv2_detection_model(cv2_org_frame: numpy.array) pytb.output.bboxes_2d.BBoxes2D [source]
Performs a YOLOv2-3-4 inference on the given frame using cv2-DetectionModel of openCV.
- Parameters
frame (np.array) – The frame to infer YOLOv2-3-4 detections.
- Returns
A set of 2D bounding boxes identifying the detected objects.
- Return type
- _detect_cv2_read_net(blob_org_frame) pytb.output.bboxes_2d.BBoxes2D [source]
Performs a YOLOv2-4 inference on the given frame using cv2-ReadNet of openCV.
- Parameters
frame (Any) – The frame to infer YOLOv2-3-4 detections.
- Returns
A set of 2D bounding boxes identifying the detected objects.
- Return type
yolo_5
- class YOLO5(proc_parameters: dict)[source]
Bases:
pytb.detection.bboxes.bboxes_2d_detector.bboxes_2d_detector.BBoxes2DDetector
- __init__(proc_parameters: dict)[source]
Initializes the detector with the given parameters.
- Parameters
proc_parameters (dict) – A dictionary containing the YOLO detector parameters
- detect(frame: numpy.array) pytb.output.bboxes_2d.BBoxes2D [source]
Performs a YOLOv5 inference on the given frame using ultralytics on PyTorch.
- Parameters
frame (np.array) – The frame to infer YOLOv5 detections.
- Returns
A set of 2D bounding boxes identifying the detected objects.
- Return type