Config Parameters
General
bag_path
Type: String
The path from the root of the project to the ROS bag file to be processed.
pose_number
Type: Integer
The number of lidar scans (and poses) to load from the bag file.
odom_topic
Type: String
The topic to pull odometry data from, this topic should have the nav_msgs/Odometry data type, or if differential_odom is set to true, it can be of type CobotMsgs/CobotOdometry.
lidar_topic
Type: String
The topic to pull LiDAR data from, this topic should have the sensor_msgs/LaserScan type.
differential_odom
Type: Boolean
True if using CobotOdometry message, false otherwise.
pose_output_file
Type: String
The path to the file to output the pose positions too, this writing will be triggered when the user hits the WriteButton in RViz.
map_output_file
Type: String
The path to output the text file to output the vector map to. If unset then when the user hits the Vectorize Button in RViz there will be no text output.
Performance
translation_weight
Type: Double
The multiplier applied to the translation error when optimizing the pose-graph.
rotation_weight
Type: Double
The multiplier applied to the rotation error when optimizing the pose-graph.
stopping_accuracy
Type: Double
When the scans change by less than this amount (in meters), the optimization iteration is deemed solved.
max_lidar_range
Type: Double
The maximum distance LiDAR data is allowed to be from the robot.
rotation_change_for_lidar
Type: Double
The amount of rotation needed to signal a lidar scan capture (in radians).
translation_change_for_lidar
Type: Double
The amount of translation needed to signal a lidar scan capture.
lidar_constraint_amount
Type: Integer
The number of poses to compare each pose too. For example the default is 10, therefore we compare and try to minimize error between the current pose and the past ten, for every pose.
outlier_threshold
Type: Double
Points further than this distance (in meters) cannot be counted as the same point during ICL/ICP in the optimization phase.
Human-In-The-Loop
hitl_lc_topic
Type: String
The topic that HITL Slam tool send the message to trigger HITL Slam, probably best to not change unless you know what you are doing.
hitl_line_width
Type: Double
Points further than this will not count as falling on the HITL LC line and therefore will not be used for HITLSlam.
hitl_pose_point_threshold
Type: Integer
The amount of points needed to include a pose for HITL Slam.
Autonomous Loop Closure
auto_lc
Type: Boolean
Automatically loop close or not
keyframe_chi_squared_test
Type: Boolean
Whether or not to use chi_squared test for keyframes
keyframe_min_odom_distance
Type: Double
Distance between keyframes if chi^2 is not in use
keyframe_local_uncertainty_filtering
Type: Boolean
Whether or not to use local uncertainty filtering for keyframes
local_uncertainty_condition_threshold
Type: Double
All scans with local uncertainty less than this threshold are one step closer to being used for automatic lc, if keyframe_local_uncertainty_filtering is on
local_uncertainty_scale_threshold
Type: Double
All scans with local uncertainty scale less than this threshold are one step closer to being used for automatic lc, if keyframe_local_uncertainty_filtering is on
local_uncertainty_prev_scans
Type: Integer
The amount of previous scans to use for calculating local uncertainty if keyframe_local_uncertainty_filter ing is on
lc_match_threshold
Type: Double
Threshold used in automatic loop closure.
lc_base_max_range
Type: Double
Base max range to consider a loop closure
lc_max_range_scaling
Type: Double
How much max range to consider a loop closure increases as nodes get more distant
lc_translation_weight
Type: Integer
Loop closure translation multiplier, used during loop closure for odometry residuals.
lc_rotation_weight
Type: Double
Loop closure rotation multiplier, used during loop closure for odometry residuals.
lc_min_keyframes
Type: Integer
Minimum number of keyframes that must exist between loop closures.
lc_debug_output_dir
Type: String
Used to dump images from auto-lc
Normal Computation
nc_neighborhood_size
Type: Double
The neighborhood size to consider when RANSACing for normals.
nc_neighborhood_step_size
Type: Double
How much the neighborhood increases with each iteration.
nc_mean_distance
Type: Double
You got me, this one's just a constant.
nc_bin_number
Type: Integer
The number of buckets to use in the Hough transform.