class NdProfile():
def __init__(self):
rospy.init_node('pub_profile3d', anonymous=True)
image_topic = rospy.get_param('~image', '/camera/image')
cloud_topic = rospy.get_param('~cloud', '/camera/cloud')
config_file = rospy.get_param('~config', 'profile3d.yaml')
#peaks_topic = rospy.get_param('~peaks', '/camera/peaks')
#self.image_pub = rospy.Publisher(peaks_topic, Image, queue_size=10)
rospy.Subscriber(
image_topic, Image, self.sub_image_topic, queue_size=1)
self.sequence = 0
self.bridge = CvBridge()
self.pcloud = PointCloud2()
self.cloud_pub = rospy.Publisher(
cloud_topic, PointCloud2, queue_size=10)
self.profile = Profile()
self.profile.load_configuration(config_file)
rospy.spin()
def pub_point_cloud(self, stamp, profile3d):
self.sequence = self.sequence + 1
self.pcloud = pc2.create_cloud_xyz32(self.pcloud.header, profile3d)
self.pcloud.header = std_msgs.msg.Header(frame_id="/camera0",
stamp=stamp,
seq=self.sequence)
self.cloud_pub.publish(self.pcloud)
def image_pub_peak(self, stamp, image):
image_msg = self.bridge.cv2_to_imgmsg(image, encoding='bgr8')
image_msg.header.stamp = stamp
self.image_pub.publish(image_msg)
def sub_image_topic(self, data):
try:
stamp = rospy.Time.now()
image = self.bridge.imgmsg_to_cv2(data)
#rospy.loginfo(stamp)
#stamp = data.header.stamp
#rospy.loginfo(stamp)
profile3d, profile2d = self.profile.points_profile(image)
if len(profile3d) > 0:
self.pub_point_cloud(stamp, profile3d)
#print profile3d
#if data.encoding == 'mono8':
# image = cv2.cvtColor(image, cv2.COLOR_GRAY2BGR)
# if len(profile3d) > 0:
# image = self.profile.draw_points(image, profile2d,
# color=(0, 0, 255),
# thickness=2)
#self.image_pub_peak(stamp, image)
except CvBridgeError, e:
print e
def __init__(self):
rospy.init_node('pub_profile3d', anonymous=True)
image_topic = rospy.get_param('~image', '/camera/image')
cloud_topic = rospy.get_param('~cloud', '/camera/cloud')
config_file = rospy.get_param('~config', 'profile3d.yaml')
#peaks_topic = rospy.get_param('~peaks', '/camera/peaks')
#self.image_pub = rospy.Publisher(peaks_topic, Image, queue_size=10)
rospy.Subscriber(image_topic,
Image,
self.sub_image_topic,
queue_size=1)
self.sequence = 0
self.bridge = CvBridge()
self.pcloud = PointCloud2()
self.cloud_pub = rospy.Publisher(cloud_topic,
PointCloud2,
queue_size=10)
self.profile = Profile()
self.profile.load_configuration(config_file)
rospy.spin()
class NdProfile():
def __init__(self):
rospy.init_node('pub_profile3d', anonymous=True)
image_topic = rospy.get_param('~image', '/camera/image')
cloud_topic = rospy.get_param('~cloud', '/camera/cloud')
config_file = rospy.get_param('~config', 'profile3d.yaml')
#peaks_topic = rospy.get_param('~peaks', '/camera/peaks')
#self.image_pub = rospy.Publisher(peaks_topic, Image, queue_size=10)
rospy.Subscriber(image_topic,
Image,
self.sub_image_topic,
queue_size=1)
self.sequence = 0
self.bridge = CvBridge()
self.pcloud = PointCloud2()
self.cloud_pub = rospy.Publisher(cloud_topic,
PointCloud2,
queue_size=10)
self.profile = Profile()
self.profile.load_configuration(config_file)
rospy.spin()
def pub_point_cloud(self, stamp, profile3d):
self.sequence = self.sequence + 1
self.pcloud = pc2.create_cloud_xyz32(self.pcloud.header, profile3d)
self.pcloud.header = std_msgs.msg.Header(frame_id="/camera0",
stamp=stamp,
seq=self.sequence)
self.cloud_pub.publish(self.pcloud)
def image_pub_peak(self, stamp, image):
image_msg = self.bridge.cv2_to_imgmsg(image, encoding='bgr8')
image_msg.header.stamp = stamp
self.image_pub.publish(image_msg)
def sub_image_topic(self, data):
try:
stamp = rospy.Time.now()
image = self.bridge.imgmsg_to_cv2(data)
#rospy.loginfo(stamp)
#stamp = data.header.stamp
#rospy.loginfo(stamp)
profile3d, profile2d = self.profile.points_profile(image)
if len(profile3d) > 0:
self.pub_point_cloud(stamp, profile3d)
except CvBridgeError, e:
rospy.loginfo("CvBridge Exception")
def __init__(self):
rospy.init_node('viewer', anonymous=True)
image_topic = rospy.get_param('~image', '/ueye/image')
# Pattern parameters
pattern_rows = rospy.get_param('~pattern_rows', 7)
pattern_cols = rospy.get_param('~pattern_cols', 8)
pattern_size = rospy.get_param('~pattern_size', 10)
config_file = rospy.get_param('~config', 'profile3d.yaml')
rospy.Subscriber(image_topic, Image, self.callback, queue_size=10)
rospy.on_shutdown(self.on_shutdown_hook)
self.counter = 0
self.bridge = CvBridge()
self.listener = tf.TransformListener()
self.square_size = pattern_size
self.grid_size = (pattern_cols - 1, pattern_rows - 1)
self.laser_profile = Profile(axis=1, thr=180, method='pcog')
self.camera_calibration = CameraCalibration(
grid_size=self.grid_size, square_size=self.square_size)
cv2.namedWindow('viewer')
cv2.cv.SetMouseCallback('viewer', self.on_mouse, '')
rospy.spin()
def __init__(self):
rospy.init_node('pub_profile3d', anonymous=True)
image_topic = rospy.get_param('~image', '/camera/image')
cloud_topic = rospy.get_param('~cloud', '/camera/cloud')
config_file = rospy.get_param('~config', 'profile3d.yaml')
#peaks_topic = rospy.get_param('~peaks', '/camera/peaks')
#self.image_pub = rospy.Publisher(peaks_topic, Image, queue_size=5)
rospy.Subscriber(image_topic, Image, self.sub_image_topic,
queue_size=1)
self.sequence = 0
self.bridge = CvBridge()
self.pcloud = PointCloud2()
self.cloud_pub = rospy.Publisher(cloud_topic, PointCloud2,
queue_size=5)
self.profile = Profile()
path = rospkg.RosPack().get_path('etna_scanning')
self.profile.load_configuration(os.path.join(path, 'config',
config_file))
rospy.spin()
np.set_printoptions(precision=4, suppress=True)
path = '../'
pattern_rows = 7
pattern_cols = 8
pattern_size = 0.010
config_file = 'profile3d.yaml'
dirname = '../data'
images, tool_poses = read_calibration_data(dirname)
square_size = pattern_size
grid_size = (pattern_cols - 1, pattern_rows - 1)
laser_profile = Profile(axis=1, thr=180, method='pcog')
camera_calibration = CameraCalibration(grid_size=grid_size,
square_size=square_size)
laser_calibration = LaserCalibration(grid_size=grid_size,
square_size=square_size,
profile=laser_profile)
laser_calibration.find_calibration_3d(images)
laser_calibration.save_parameters(os.path.join(path, 'config', config_file))
filenames = sorted(glob.glob('../data/pose*.txt'))
ks = [int(filename[-8:-4]) for filename in filenames]
poses_checker, poses_tool = [], []
poses_ichecker, poses_itool = [], []
for k in ks:
print 'Frame: %i' % k
img = read_image('../data/frame%04i.png' % k)
def __init__(self, grid_size=(7, 6), square_size=10.0, profile=Profile()):
CameraCalibration.__init__(self,
grid_size=grid_size,
square_size=square_size)
self.camera_pose = (np.eye(3), np.zeros(3))
self.profile = profile
def read_calibration_data(dirname):
frame_filenames = sorted(glob.glob(os.path.join(dirname, 'frame0*.png')))
pose_filenames = sorted(glob.glob(os.path.join(dirname, 'pose0*.txt')))
images = [read_image(filename) for filename in frame_filenames]
tool_poses = [read_pose(filename) for filename in pose_filenames]
return images, tool_poses
if __name__ == '__main__':
from mlabplot import MPlot3D
dirname = '../../data'
images, tool_poses = read_calibration_data(dirname)
laser_profile = Profile(axis=1, thr=180, method='pcog')
laser_calibration = LaserCalibration(grid_size=(7, 6),
square_size=0.010,
profile=laser_profile)
laser_calibration.find_calibration_3d(images)
pattern_poses = laser_calibration.pattern_poses
profiles = [laser_profile.profile_points(img) for img in images]
lines = [
laser_calibration.find_best_line2d(profile2d) for profile2d in profiles
]
for k, img in enumerate(images):
grid = laser_calibration.grids[k]
imgc = laser_calibration.draw_chessboard(img.copy(), grid)
pose = laser_calibration.get_chessboard_pose(grid)
def read_calibration_data(dirname):
frame_filenames = sorted(glob.glob(os.path.join(dirname, 'frame0*.png')))
pose_filenames = sorted(glob.glob(os.path.join(dirname, 'pose0*.txt')))
images = [read_image(filename) for filename in frame_filenames]
tool_poses = [read_pose(filename) for filename in pose_filenames]
return images, tool_poses
if __name__ == '__main__':
from mlabplot import MPlot3D
dirname = '../../data'
images, tool_poses = read_calibration_data(dirname)
laser_profile = Profile(axis=1, thr=180, method='pcog')
laser_calibration = LaserCalibration(grid_size=(7, 6),
square_size=0.010,
profile=laser_profile)
laser_calibration.find_calibration_3d(images)
pattern_poses = laser_calibration.pattern_poses
profiles = [laser_profile.profile_points(img)
for img in images]
lines = [laser_calibration.find_best_line2d(profile2d)
for profile2d in profiles]
for k, img in enumerate(images):
grid = laser_calibration.grids[k]
imgc = laser_calibration.draw_chessboard(img.copy(), grid)
pose = laser_calibration.get_chessboard_pose(grid)