def cb_X0(f):
global cTsAry,bTmAry
print "cbX0"
pb_msg.publish("rcalib::clear")
cTsAry=TransformArray()
bTmAry=TransformArray()
count=Int32(); count.data=len(cTsAry.transforms); pb_count.publish(count)
def __init__(self):
# Frame which is rigidly attached to the camera
# The transform from this frame to the camera optical frame is what
# we're trying to compute.
# For the eye-in-hand case, this is the end-effector frame.
# For the eye-on-base case, this is the world or base frame.
self.camera_parent_frame_id = rospy.get_param('~camera_parent_frame')
# Frame which is rigidly attached to the marker
# The transform from the camera parent frame to the marker parent frame
# is given by forward kinematics.
# For the eye-in-hand case, this is the world or base frame.
# For the eye-on-base case, this is the end-effector frame.
self.marker_parent_frame_id = rospy.get_param('~marker_parent_frame')
self.publish_tf = rospy.get_param('~publish_tf')
self.tf_suffix = rospy.get_param('~tf_suffix')
self.marker_id = rospy.get_param('~marker_id')
self.optical_frame_id = 'cc'
# Compute the camera base to optical transform
self.xyz_optical_base = rospy.get_param('~xyz_optical_base', [0, 0, 0])
self.rpy_optical_base = rospy.get_param('~rpy_optical_base', [0, 0, 0])
self.F_optical_base = PyKDL.Frame(
PyKDL.Rotation.RPY(*self.rpy_optical_base),
PyKDL.Vector(*self.xyz_optical_base))
self.F_base_optical = self.F_optical_base.Inverse()
# tf structures
self.listener = tf.TransformListener()
self.broadcaster = tf.TransformBroadcaster()
# input data
self.hand_world_samples = TransformArray()
self.camera_marker_samples = TransformArray()
# marker subscriber
# self.aruco_subscriber = rospy.Subscriber(
# 'aruco_tracker/transform',
# TransformStamped,
# self.aruco_cb,
# queue_size=1)
# calibration service
self.hand_eye_service = rospy.Service('hand_eye_calibration',
hand_eye_calibration,
self.aruco_cb)
self.hand_eye_service = rospy.Service('compute_calibration',
hand_eye_calibration,
self.compute_calibration)
rospy.wait_for_service('compute_effector_camera_quick')
self.calibrate = rospy.ServiceProxy('compute_effector_camera_quick',
compute_effector_camera_quick)
self.samples_min = 5
def get_visp_samples(self):
hand_world_samples = TransformArray()
hand_world_samples.header.frame_id = self.tracking_base_frame # bug?
camera_marker_samples = TransformArray()
camera_marker_samples.header.frame_id = self.tracking_base_frame
for s in self.samples:
camera_marker_samples.transforms.append(s['optical'])
hand_world_samples.transforms.append(s['robot'])
return hand_world_samples, camera_marker_samples
def __init__(self):
# Frame which is rigidly attached to the camera
# The transform from this frame to the camera optical frame is what
# we're trying to compute.
# For the eye-in-hand case, this is the end-effector frame.
# For the eye-on-base case, this is the world or base frame.
self.camera_parent_frame_id = rospy.get_param('~camera_parent_frame')
# Frame which is rigidly attached to the marker
# The transform from the camera parent frame to the marker parent frame
# is given by forward kinematics.
# For the eye-in-hand case, this is the world or base frame.
# For the eye-on-base case, this is the end-effector frame.
self.marker_parent_frame_id = rospy.get_param('~marker_parent_frame')
self.publish_tf = rospy.get_param('~publish_tf')
self.tf_suffix = rospy.get_param('~tf_suffix')
self.sample_rate = rospy.get_param('~sample_rate')
self.interactive = rospy.get_param('~interactive')
# Compute the camera base to optical transform
self.xyz_optical_base = rospy.get_param('~xyz_optical_base', [0, 0, 0])
self.rpy_optical_base = rospy.get_param('~rpy_optical_base', [0, 0, 0])
self.F_optical_base = PyKDL.Frame(
PyKDL.Rotation.RPY(*self.rpy_optical_base),
PyKDL.Vector(*self.xyz_optical_base))
self.F_base_optical = self.F_optical_base.Inverse()
# tf structures
self.listener = tf.TransformListener()
self.broadcaster = tf.TransformBroadcaster()
# rate limiter
self.rate = rospy.Rate(self.sample_rate)
# input data
self.hand_world_samples = TransformArray()
self.camera_marker_samples = TransformArray()
# marker subscriber
self.QR_subscriber = rospy.Subscriber(
'visp_auto_tracker/object_position',
PoseStamped,
self.pose_cb,
queue_size=1)
rospy.loginfo('Subscribe to the QR object pose topic!')
# calibration service
rospy.wait_for_service('compute_effector_camera_quick')
self.calibrate = rospy.ServiceProxy('compute_effector_camera_quick',
compute_effector_camera_quick)
rospy.loginfo('Calibration service available!')
def cb_X3(f):
global cTsAry, bTmAry
print "X3"
Tcsv = np.loadtxt('input.txt')
bTmAry = TransformArray()
cTsAry = TransformArray()
for vec in Tcsv:
# bTmAry.transforms.append(xyz2quat(tflib.fromVec(vec[0:7])))
bTmAry.transforms.append(tflib.fromVec(vec[0:7]))
cTsAry.transforms.append(tflib.fromVec(vec[7:14]))
print bTmAry.transforms[0]
print cTsAry.transforms[0]
call_visp()
save_result_mTs('result_mts.txt')
save_result_bTs('result_bts.txt')
return
def get_visp_samples(self):
"""
Returns the sample list as a TransformArray.
:rtype: visp_hand2eye_calibration.msg.TransformArray
"""
hand_world_samples = TransformArray()
# hand_world_samples.header.frame_id = self.robot_base_frame
hand_world_samples.header.frame_id = self.tracking_base_frame
# DONTFIXME: yes, I know, it should be like the line above.
# thing is, otherwise the results of the calibration are wrong. don't ask me why
camera_marker_samples = TransformArray()
camera_marker_samples.header.frame_id = self.tracking_base_frame
for s in self.samples:
camera_marker_samples.transforms.append(s['optical'].transform)
hand_world_samples.transforms.append(s['robot'].transform)
return hand_world_samples, camera_marker_samples
def __init__(self, name):
self.name = name
# the names of the base_link and end_link according to the urdf model
self.base_link = "iiwa_base_link"
self.end_link = "iiwa_adapter"
# two variables to compute the time fragment between the two callbacks
self.time_A = dt.datetime.now()
self.time_B = dt.datetime.now()
self.joint_names = [
"joint1", "joint2", "joint3", "joint4", "joint5", "joint6",
"joint7"
]
# initialize a node
rospy.init_node("path_walker_node")
self.robot = URDF.from_xml_file(
"/home/davoud/kuka_ws/src/kuka/kuka/iiwa/iiwa_description/urdf/iiwa.urdf"
)
self.tree = kdl_tree_from_urdf_model(self.robot)
self.kdl_kin = KDLKinematics(self.robot, self.base_link, self.end_link)
# open the calibration_data file for saving the coordinates of end_effector and marker
self.file = open("calibration_data", 'a')
self.file.write("\n\nNew Run Time: " + str(dt.datetime.now()) +
"\n")
# current position of the end effector
# self.end_eff_current_pose = np.empty((1, 7))
# number of poses for the path
self.num_of_poses = 13
# current pose of path
self.current_pose = 0
# the two transform arrays to call the service
self.world_effector_transform_array = TransformArray()
self.world_effector_transform_array.header.stamp = rospy.Time.now()
self.world_effector_transform_array.header.frame_id = "iiwa_base_link"
self.camera_object_transform_array = TransformArray()
self.camera_object_transform_array.header.stamp = self.world_effector_transform_array.header.stamp
self.camera_object_transform_array.header.frame_id = "camera_rgb_optical_frame"
# list of poses
"""
self.pose1 = [-0.38, -1.79, 0.00, 0.00, -1.48, -1.90, -1.32]
self.pose3 = [-0.65, -1.51, 0.13, 0.00, -1.44, -1.75, -1.06]
self.pose5 = [-0.81, -1.33, 0.17, -0.37, -1.25, -1.71, -0.71]
self.pose7 = [-0.45, -1.33, 0.17, -0.79, -1.42, -1.76, -0.40]
self.pose9 = [-0.06, -0.90, 0.17, -1.28, -1.88, -1.76, -0.40]
self.pose11 = [-0.06, -0.20, 0.17, -1.28, -1.88, -1.76, -0.42]
self.pose13 = [-0.06, 0.27, 0.16, -1.06, -1.88, -1.76, -0.08]
self.pose15 = [-0.05, 1.25, 0.15, -0.43, -1.88, -1.76, -0.08]
self.pose17 = [-1.64, 1.21, 0.97, -1.80, -2.95, -1.57, -2.20]
self.pose19 = [ 0.10, 1.09, 0.97, -0.19, -2.94, -1.57, -1.83]
self.pose21 = [ 0.78, 1.10, -1.08, -0.19, -0.98, -1.33, -1.51]
self.pose23 = [ 1.54, 1.34, -1.08, -0.19, -2.15, -0.66, -1.30]
self.pose25 = [ 1.98, 1.13, -1.27, -0.68, -2.15, -0.66, -1.30]
self.pose27 = [ 2.15, 1.93, -2.17, -0.48, -2.45, -1.06, -1.10]
self.pose29 = [ 2.15, 1.63, -2.17, -0.48, -2.15, -0.86, -1.30]
self.pose31 = [ 0.85, -0.23, -2.77, 1.88, -2.15, 0.46, 0.90]
self.pose2 = [-0.38, -1.79, 0.00, 0.00, -1.48, -1.70, -1.42]
self.pose4 = [-0.65, -1.51, 0.13, 0.00, -1.44, -1.55, -1.16]
self.pose6 = [-0.81, -1.33, 0.17, -0.37, -1.25, -1.51, -0.81]
self.pose8 = [-0.45, -1.33, 0.17, -0.79, -1.42, -1.56, -0.50]
self.pose10 = [-0.06, -0.90, 0.17, -1.28, -1.88, -1.56, -0.50]
self.pose12 = [-0.06, -0.20, 0.17, -1.28, -1.88, -1.56, -0.52]
self.pose14 = [-0.06, 0.27, 0.16, -1.06, -1.88, -1.56, -0.18]
self.pose16 = [-0.05, 1.25, 0.15, -0.43, -1.88, -1.56, -0.18]
self.pose18 = [-1.64, 1.21, 0.97, -1.80, -2.95, -1.37, -2.30]
self.pose20 = [ 0.10, 1.09, 0.97, -0.19, -2.94, -1.77, -1.93]
self.pose22 = [ 0.78, 1.10, -1.08, -0.19, -0.98, -1.53, -1.61]
self.pose24 = [ 1.54, 1.34, -1.08, -0.19, -2.15, -0.56, -1.40]
self.pose26 = [ 1.98, 1.13, -1.27, -0.68, -2.15, -0.56, -1.40]
self.pose28 = [ 2.15, 1.93, -2.17, -0.48, -2.45, -1.06, -1.20]
self.pose30 = [ 2.15, 1.63, -2.17, -0.48, -2.15, -0.86, -1.00]
self.pose32 = [ 0.85, -0.23, -2.77, 1.88, -2.15, 0.26, 0.80]
"""
self.pose1 = [0.39, 1.67, -2.77, 0.07, -2.15, 0.26, 1.19]
self.pose2 = [0.44, 1.67, -2.77, 0.07, -2.15, 0.26, 1.19]
self.pose3 = [0.49, 1.67, -2.77, 0.07, -2.15, 0.26, 1.19]
self.pose4 = [0.34, 1.67, -2.77, 0.07, -2.15, 0.26, 1.19]
self.pose5 = [0.29, 1.67, -2.77, 0.07, -2.15, 0.26, 1.19]
self.pose6 = [0.39, 1.72, -2.77, 0.07, -2.15, 0.26, 1.19]
self.pose7 = [0.39, 1.77, -2.77, 0.07, -2.15, 0.26, 1.19]
self.pose8 = [0.39, 1.62, -2.77, 0.07, -2.15, 0.26, 1.19]
self.pose9 = [0.39, 1.57, -2.77, 0.07, -2.15, 0.26, 1.19]
self.pose10 = [0.39, 1.57, -2.77, 0.07, -2.15, 0.26, 1.14]
self.pose11 = [0.39, 1.57, -2.77, 0.07, -2.15, 0.26, 1.09]
self.pose12 = [0.39, 1.57, -2.77, 0.07, -2.15, 0.26, 1.24]
self.pose13 = [0.39, 1.57, -2.77, 0.07, -2.15, 0.26, 1.29]
self.pose14 = [0.39, 1.57, -2.77, 0.07, -2.15, 0.26, 1.34]
self.pose15 = [0.39, 1.57, -2.77, 0.07, -2.15, 0.26, 1.39]
self.pose16 = [0.39, 1.57, -2.77, 0.07, -2.15, 0.26, 1.44]
self.pose17 = [0.39, 1.57, -2.77, 0.07, -2.15, 0.26, 1.49]
# transform listener
self.transform_listener = tf.TransformListener()
# the two current transforms
self.world_effector_transform = Transform()
self.camera_object_transform = Transform()
# actionlib preparation
self.manipulator_client = actionlib.SimpleActionClient(
"/iiwa/follow_joint_trajectory", FollowJointTrajectoryAction)
self.manipulator_client.wait_for_server()
self.trajectory = JointTrajectory()
self.trajectory.joint_names = self.joint_names
self.trajectory.points.append(JointTrajectoryPoint())
# subscribe to "/joint_states" topic which is published by kuka.
rospy.Subscriber("/joint_states",
JointState,
self.kuka_inf_cb,
queue_size=1)
# subscribe to "/aruco_single/marker" which is published by aruco_ros
# rospy.Subscriber("aruco_single/marker", Marker, self.marker_pose_cb)
# subscribe to to "/avg_marker" which is computed by averaging two markers
rospy.Subscriber("/avg_marker", Marker, self.marker_pose_cb)
print 'Visp reset failed:' + e
pb_msg.publish("rcalib::visp reset failed")
return
calibrator = None
try: #solving as fixed camera
calibrator = rospy.ServiceProxy('/compute_effector_camera_quick',
compute_effector_camera_quick)
except rospy.ServiceException, e:
print 'Visp call failed:' + e
pb_msg.publish("rcalib::visp call failed")
return
creset(resetRequest())
req = compute_effector_camera_quickRequest()
req.camera_object = cTsAry
if Config["mount_frame_id"] == "world": #fixed camera
mTbAry = TransformArray()
for tf in bTmAry.transforms:
mTbAry.transforms.append(tflib.inv(tf))
req.world_effector = mTbAry
print "calib fixed"
else:
req.world_effector = bTmAry
print "calib handeye"
try:
res = calibrator(req)
print res.effector_camera
set_param_tf(
Config["config_tf"] + "/" + Config["camera_frame_id"] +
"/transform", res.effector_camera)
mTc = tflib.toRT(res.effector_camera)
pb_msg.publish("rcalib::visp solver success")
def __init__(self, config):
self.config = config
self.target = None # Initialize tracked target pose
self.cam_T_target = None
self.w_T_hand = None
# Msgs for visp_hand2eye_calibration
self.cam_T_target_array = TransformArray()
self.w_T_hand_array = TransformArray()
# Publishers for visp_hand2eye_calibration
self.w_T_hand_pub = rospy.Publisher("/world_effector",
Transform,
queue_size=1000)
self.cam_T_target_pub = rospy.Publisher("/camera_object",
Transform,
queue_size=1000)
# Initialize robot
print "Loading URDF..."
self.robot = pydrake.rbtree.RigidBodyTree(self.config["ik_urdf"])
print
print "Bodies"
print "------"
self.bodies = dict()
for body in range(self.robot.get_num_bodies()):
body_name = self.robot.getBodyOrFrameName(body)
print body_name, body
self.bodies[body_name] = body
# if robot.get_num_bodies() == 0: print "No bodies"
print
self.positions = dict()
self.joint_names = list()
print "Positions"
print "---------"
for position in range(self.robot.get_num_positions()):
position_name = self.robot.get_position_name(position)
print position_name, position
self.positions[position_name] = position
self.joint_names.append(position_name)
# if robot.get_num_positions() == 0: print "No positions"
print
# Set up ROS
print "Waiting for getRobotConf service..."
rospy.wait_for_service('/robot_server/getRobotConf')
self.conf_service = rospy.ServiceProxy('/robot_server/getRobotConf',
GetRobotConf)
print "Waiting for MoveHead service..."
rospy.wait_for_service('/robot_control/MoveHead')
self.move_head_service = rospy.ServiceProxy('/robot_control/MoveHead',
MoveHead)
print "Initializing publishers and subscribers..."
self.tf_buffer = tf2_ros.Buffer(
rospy.Duration(10.0)) #tf buffer length
self.tf_listener = tf2_ros.TransformListener(self.tf_buffer)
self.tf_broadcaster = tf2_ros.TransformBroadcaster()
print "ROS stack ready!"
print
