Skip to content

Latest commit

 

History

History
116 lines (87 loc) · 5.29 KB

DOCUMENTATION.md

File metadata and controls

116 lines (87 loc) · 5.29 KB
Raw

ROS Nodes Documentation

FaceRecognition.py

Purpose: Uses the face_recognition library to detect and recognize faces in images. It subscribes to the camera topic and publishes the detected person's name and coordinates.

Input Topics:

  • /zed2/zed_node/rgb/image_rect_color (sensor_msgs/Image): The input image from the camera.

Output Topics:

  • /vision/person_detected_name (std_msgs/String): The name of the detected person.
  • /vision/person_list (frida_vision_interfaces/PersonList): The list of detected persons.
  • /vision/target (geometry_msgs/Point): The coordinates of the detected person.
  • /vision/img_face_recognition (sensor_msgs/Image): The annotated image with face recognition results.

Services:

  • /vision/new_name (frida_vision_interfaces/NewHost): Service to add a new person to the face recognition system.

Parameters:

  • MAX_DEGREE (int): The maximum degree for face tracking.
  • TRACK_THRESHOLD (int): The threshold for tracking faces.
  • AREA_THRESHOLD (int): The threshold for face area.

ReceptionistCommands.py

Purpose: Uses YOLO models to detect people and objects in images. It provides services to check for the presence of people and find available seats.

Input Topics:

  • /zed2/zed_node/rgb/image_rect_color (sensor_msgs/Image): The input image from the camera.

Output Topics:

  • /vision/img_person_detection (sensor_msgs/Image): The annotated image with person detection results.

Services:

  • /vision/check_person (std_srvs/SetBool): Service to check if a person is detected.
  • /vision/find_seat (frida_vision_interfaces/FindSeat): Service to find an available seat.

Parameters:

  • MODEL_LOCATION (str): The path to the YOLO model.
  • PERCENTAGE (float): The percentage of the image to consider for person detection.
  • MAX_DEGREE (int): The maximum degree for seat finding.

PersonTracking.py

Purpose: Uses YOLO and ReID models to track people in images. It publishes the coordinates of the tracked person.

Input Topics:

  • /zed2/zed_node/rgb/image_rect_color (sensor_msgs/Image): The input image from the camera.

Output Topics:

  • /vision/person_detection (geometry_msgs/Point): The coordinates of the tracked person.
  • /vision/img_tracking (sensor_msgs/Image): The annotated image with person tracking results.

Services:

  • /vision/change_person_tracker_state (std_srvs/SetBool): Service to enable or disable person tracking.

Parameters:

  • FLIP_IMAGE (bool): Whether to flip the input image.

bagDetector.py

Purpose: Uses YOLO models to detect bags in images. It publishes the detected bags and their 3D coordinates.

Input Topics:

  • /zed2/zed_node/rgb/image_rect_color (sensor_msgs/Image): The input image from the camera.
  • /camera/depth/image_raw (sensor_msgs/Image): The input depth image from the camera.
  • /camera/depth/camera_info (sensor_msgs/CameraInfo): The camera info for the depth image.

Output Topics:

  • detections/bag (bag_detector/objectDetectionArray): The detected bags.
  • detections_image/bag (sensor_msgs/Image): The annotated image with bag detection results.
  • detections_3d/bag (visualization_msgs/MarkerArray): The 3D coordinates of the detected bags.

Parameters:

  • SOURCE (str): The source of the input image.
  • ROS_INPUT (bool): Whether to use ROS input.
  • USE_ACTIVE_FLAG (bool): Whether to use the active flag.
  • DEPTH_ACTIVE (bool): Whether to use the depth image.
  • DEPTH_INPUT (str): The input depth image topic.
  • CAMERA_INFO (str): The camera info topic.
  • MIN_SCORE_THRESH (float): The minimum score threshold for bag detection.
  • VERBOSE (bool): Whether to enable verbose output.
  • CAMERA_FRAME (str): The camera frame.
  • YOLO_BAG_MODEL_PATH (str): The path to the YOLO bag model.
  • FLIP_IMAGE (bool): Whether to flip the input image.

detectPointingObject.py

Purpose: Uses Mediapipe to detect the pointing direction of a user's hand. It provides an action server to detect the pointed object.

Input Topics:

  • /zed2/zed_node/rgb/image_rect_color (sensor_msgs/Image): The input image from the camera.
  • detections/bag (bag_detector/objectDetectionArray): The detected bags.

Output Topics:

  • pointing_direction (sensor_msgs/Image): The annotated image with pointing direction results.
  • pointed_object_marker (visualization_msgs/Marker): The marker for the pointed object.

Services:

  • detectPointingObject (frida_vision_interfaces/DetectPointingObjectAction): Action server to detect the pointed object.

Parameters:

  • USE_RIGHT_HAND (bool): Whether to use the right hand for pointing detection.
  • USE_LEFT_HAND (bool): Whether to use the left hand for pointing detection.
  • INFERENCE_TIMEOUT (int): The timeout for inference.
  • SOURCE (str): The source of the input image.
  • ROS_INPUT (bool): Whether to use ROS input.
  • USE_ACTIVE_FLAG (bool): Whether to use the active flag.
  • DEPTH_ACTIVE (bool): Whether to use the depth image.
  • DEPTH_INPUT (str): The input depth image topic.
  • CAMERA_INFO (str): The camera info topic.
  • MIN_SCORE_THRESH (float): The minimum score threshold for pointing detection.
  • VERBOSE (bool): Whether to enable verbose output.
  • CAMERA_FRAME (str): The camera frame.
  • YOLO_BAG_MODEL_PATH (str): The path to the YOLO bag model.
  • FLIP_IMAGE (bool): Whether to flip the input image.