Lecture 7: Localization Flashcards
What is Map-based localization?
The robot estimates its position using perceived information and a map
What about the 2 things of the map used in map-based localization?
It might be known.
It might be built in parallel(simultaneous localization and mapping-SLAM)
What are the 2 challenges of localization?
1)Measurements and the map are inherently error prone
2)The robot has to deal with uncertain information
What is the approach of Localization?
The robot estimates the belief state about its position through an SEE and ACT cycle
Measurements are error prone because of the following 3 things:
Odometry
Exteroceptive sensors(camera, laser)
Map
The first See involves what?
The robot queries its sensors -> finds itself next to a pillar
The Act involves what?
Robot moves one meter forward
* Motion estimated by wheel encoders
* Accumulation of uncertainty
The second See involves what?
The robot queries its sensors again -> finds itself next to a pillar
What does the belief update involve?
Information Fusion
What are the 4 types of maps?
-Continuous map with single hypothesis
probability distribution π(π₯)
-Continuous map with multiple hypotheses
probability distribution π(π₯)
-Discretized metric map (grid π) with
probability distribution π(π)
-Discretized topological map (nodes π) with
probability distribution π(π).
What does the Bayes rule relate and how is it written?
the conditional probability p(x|y) to its inverse p(y|x).
p(x|y) = np(y|x)p(x); n = p(y)^-1 normalization factor(integral of p = 1)
Who uses the Bayes Rule and where is the Bayes rule theorem used?
The theorem is used by both Markov and Kalman-filter localization algorithms during the measurement update.
Concerning application of theorem of total probability/convolution, what two probabilities is it applied to? What is the difference?
Continuous and Discrete probabilities. Continuous probabilities has dx(subscript)t-1
What is Markov localization for?
Discretized pose representation
What is Kalman filter for?
Continuous pose representation and Gaussian error model
What is the SLAM problem?
how can a body navigate in a previously unknown environment, while constantly building and updating a map of its workspace using onboard sensors & onboard computation
In what 4 situations is SLAM necessary?
-When a robot must be truly autonomous (no human input)
* When there is no prior knowledge about the environment
* When we cannot rely exclusively on pre-placed beacons or external
positioning systems (e.g., GPS)
* When the robot needs to know where it is
How do you track the motion of a robot while it is moving?
SLAM(Simultaneous Localization and Mapping)
SLAM is the backbone of what?
the spatial awareness of a robot
The spatial awareness of a robot is what?
one of the most challenging problems in probabilistic robotics.
An unbiased map is necessary for what?
Localizing the robot; pure localization with a known map
An accurate pose estimates is necessary for?
building a map of the environment; Mapping with known robot poses.
What makes SLAM better than an unbiased map and an accurate pose estimate?
SLAM has no prior knowledge of the robotβs workspace. The robot poses have to be estimated along the way.
Where does SLAM originate from?
efforts to formalize production of topographic maps from aerial imagery
What is βphotogrammetryβ?
The practice of determining the geometric properties of objects from images
What is the goal of Photogrammetry?
align the images to build a topographic map of the area
How does traditional SLAM work?
Pick natural scene features as landmarks, observe their motion & reason about robot motion.
What does traditional SLAM research into?
Good features to track, sensors, trackers, representations, assumptions
Ways of dealing with uncertainty in the processes involved
What are the 3 approaches to SLAM?
Bundle adjustment
Filtering(UKF/EKF/Particle Filter SLAM)
Keyframes
What is the Vision from SLAM?
Images = information-rich snapshots of a scene
Compactness + affordability of cameras
HW advances
When SLAM using a single,handheld camera?
Very applicable, compact, affordable
What about Structure from Motion(SFM)?
Take some images of the object/scene to reconstruct
* Features (points, lines, β¦) are extracted from all frames and matched among
them.
* Process all images simultaneously: images do not need to be ordered.
* Optimization to recover both:
* Camera motion and
* 3D structure
* Up to a scale factor
* Not real-time
MonoSLAM | problem statement
How can we track the motion of a camera while it is moving? i.e., online
Extract Shi-Tomasi features & track them in image space.
SLAM using a single camera, grabbing frames at 30Hz
Ellipses (in camera view) and Ellipsoids (in map view) represent uncertainty
What about ORB-SLAM
The most powerful open-source
monocular SLAM approach today.
Uses ORB features (binary) in a
keyframe-based approach.
Binary place recognition
What about OKVIS: open keyframe-based Visual Inertial SLAM
Visual-inertial SLAM odometry approach (i.e.
no loop-closure)
* Uses BRISK features in a keyframe-based
approach
* Tight visual-inertial fusion β handles both
monocular and stereo vision
What about ROVIO?
- EKF-based
- Detects a variant of Shi-Tomasi
features at different scale levels - Tracks patches and uses the intensity
errors in the innovation term - Can only track a limited no. features,
so ROVIO performs odometry, not
SLAM.
What about VINS-mono?
- a robust and versatile monocular visual-inertial state estimator
- Keyframe-based approach
- Runs SLAM based on a tightly- coupled visual-inertial odometry with relocalization.
- Shi-Tomasi features tracked using the KLT sparse optical flow tracker
- BRIEF descriptors for relocalization, using binary place recognition
- Extensions to stereo and IMU open- sourced.
Whatβs next for SLAM
- Centralized: all data needs to pass through the server
Remove redundant data
Work towards distributed collaboration - Server accessibility limits mission range
Use cloud computing or a mobile server
Or peer-to-peer communication in a distributed architecture. - Enable stronger collaboration
