Camera

For the perception, we use Camera Chameleon, a digital USB 2.0 camera Chameleon manufactured by PointGrey.

From Python, the library used for the communication with the camera is PyCapture2. Here is an example of camera usage in Python:

import PyCapture2
import cv2
import numpy as np
 
# Initialize bus and camera
bus = PyCapture2.BusManager()
camera = PyCapture2.Camera()
 
# Select first camera on the bus
camera.connect(bus.getCameraFromIndex(0))
 
# Start capture
camera.startCapture()
 
while True:
  # Retrieve image from camara in PyCapture2.Image format
  image = camera.retrieveBuffer()
 
  # Convert from MONO8 to RGB8
  image = image.convert(PyCapture2.PIXEL_FORMAT.RGB8)
 
  # Convert image to Numpy array
  rgb_cv_image = np.array(image.getData(), dtype="uint8").reshape((image.getRows(), image.getCols(), 3));
 
  # Convert RGB image to BGR image to be shown by OpenCV
  bgr_cv_image = cv2.cvtColor(rgb_cv_image, cv2.COLOR_RGB2BGR)
 
  # Show image
  cv2.imshow('frame',bgr_cv_image)
 
  # Wait for key press, stop if the key is q
  if cv2.waitKey(1) & 0xFF == ord('q'):
    break

More documentation about PyCapture2 can be found with this command:

evince /opt/salt-install/PyCapture2-2.13.31/docs/PyCapture2doc.pdf

Driver documentation: pointgray_camera_driver

Start camera driver nodes:

# Assuming that the B3B33ROB ROS environment was properly sourced, i.e. :
source /opt/ros/b3b33rob
 
# just camera driver
roslaunch b3b33rob_ros camera.launch
 
# or camera with robot
roslaunch b3b33rob_ros start_rv6sdl.launch camera:=true
roslaunch b3b33rob_ros start_rv6s.launch camera:=true
roslaunch b3b33rob_ros start_fake_rv6sdl.launch camera:=true
roslaunch b3b33rob_ros start_fake_rv6sdl.launch camera:=true

Display camera stream:

rosrun image_view image_view image:=/camera/image_color

In order to configure the camera use the FlyCap2 utility. In terminal run:

flycap2