Target object
Sensor
Robot/Platform
Simulation environment
Spatial representation
View search space
Shape knowledge
View determination
Path planning
Termination
Performance evaluation
Structured light
Stereo camera
Monocular camera
Time-of-Flight (ToF)
Range sensor
(unspecified)
Manual operation
Robot arm
Ground mobile robot
Ground mobile robot
with robot arm
Micro aerial vehicle
(MAV)
Not considered
In-house
Gazebo
AirSim
V-REP
No simulation
2D occupancy map
3D voxel grid
3D occupancy map
with label
3D occupancy map
with probability
2D/3D vector field
3D TSDF
2D polygonal layout
3D point cloud
Planar view space
3D view space
Spherical view space
Prior 3D
pre-scanned
3D object recognition
Machine learning
Frontier-based
exploration
# of unknown voxel
Machine learning
Information-theoretic
(entropy reduction)
Seen free space
Fixed # of operation
Negligible
improvement/change
Completeness
(mesh/voxel or
feasible views)
Maximum # of scan
Not described
Visual completeness
Visual trajectory path
# of scans
Time
Distance traveled
Volume (explored)
Surface (scan
coverage)
Entropy reduced
Reconstruction error
Self measurement
only
None
Indoor
Not considered
Vector field guided
Visual information
Occlusion boundary
surface gain
Sampling density on
surface
Reinforcement
learning
Information gain in
edges/actions
RRT approach
Genetic approach
Environment
modeling
Dijkstra approach
Not considered
A* approach
Potential fields
Consider motion
uncertainty
Outside the
bounding box
Real-time obstacle
detection