vision_system.camera module

class vision_system.camera.Camera(color_image_topic: str = '/camera/color/image_raw', depth_image_topic: str = '/camera/depth/image_raw', camera_info_topic: str = '/camera/depth/camera_info', frames_approx_sync: bool = False, depth_frame_encoding: str = '32FC1')

Bases: Node

Camera class for acquiring images (color and depth) from an RGBD-camera.

This class is responsible for subscribing to the image and depth topics making it easy to access to camera frames, converting the images using cv_bridge, and optionally applying post-processing.

Parameters

color_image_topicstr, optional

The topic for color images (default is ‘/camera/color/image_raw’).

depth_image_topicstr, optional

The topic for depth images (default is ‘/camera/depth/image_raw’).

camera_info_topicstr, optional

The topic for camera information (default is ‘/camera/depth/camera_info’).

frames_approx_syncbool, optional

Whether to use approximate time synchronization (default is False).

depth_frame_encodingstr, optional

The encoding of the depth frame (default is ‘32FC1’).

acquire_color_frame_once() → ndarray | None

Acquires a single color frame waiting for the topic.

Returns:

The acquired color frame as a cv2 image as NumPy array, or None if acquisition fails.

acquire_frames_once() → Tuple[ndarray | None, ndarray | None]

Acquires a single pair of color and depth frames.

Returns:

A tuple of (color_frame, depth_frame) as NumPy arrays, or (None, None) if acquisition fails.

get_camera_info() → CameraInfo | None

Retrieves the camera info.

Returns:

The camera info as a CameraInfo object, or None if it has not been retrieved.

get_color_frame() → ndarray | None

Retrieves the most recently acquired color frame.

Returns:

The color frame as a NumPy array, or None if no frame is available.

get_distance_frame() → ndarray | None

Retrieves the most recently acquired depth frame.

Returns:

The depth frame as a NumPy array, or None if no frame is available.

get_frame_id() → str | None

Retrieves the frame ID from the camera info.

Returns:

The frame ID as a string, or None if camera info is not available.

get_frames() → Tuple[ndarray, ndarray] | None

Retrieves the most recently acquired color and depth frames.

Returns:

A tuple of (color_frame, depth_frame) as NumPy arrays, or None if frames are not available.

process_once() → ndarray | None

Processes a single pair of color and depth frames using the provided post-processing function.

Returns:

Processed frames or None if acquisition or processing fails.

retrieve_camera_info() → bool

Retrieves the camera info waiting for the topic and stores it in the class.

Returns:

True if camera info is successfully retrieved, False otherwise.

set_processing_function(package_name: str, module_name: str, class_name: str) → None

Dynamically loads and sets the post-processing function.

Parameters:

• package_name – The name of the package containing the post-processing module.

• module_name – The name of the module containing the post-processing class.

• class_name – The name of the post-processing class.

Raises:

Exception if loading fails or if the class is not an instance of PostProcessing.

start_acquire() → None

Starts the acquisition of synchronized color and depth frames, with post-processing.

start_acquire_only_color() → None

Starts the acquisition of color frames only, with post-processing (is necessary to spin the node).