Abstract:
Visual SLAM is a method where a robot moves in an environment and keeps
building map of the environment and localizing itself within the map . The
biggest problem in visual SLAM is the drift in localization that keeps on
increasing with time .
To solve this, a method known as loop closure is used where robot corrects
its map and all previous positions when it comes to a place it has already
visited . This becomes impossible when scene changes . For example, the
robot comes back to a place it has already visited but the scene, being
dynamic, has changed . The robot will not be able to close the loop . Even
the scene is dynamic, SLAM succesfully detects the scene as revisited place
using its place recognition tool. The SLAM algorithm misses the opportunity
to close the loop not only in dynamic environment but also when robot arrives
at the same scene with di erent view to the scene.
The SLAM using place recognition can be decieved in an environment
where scene is intentionally repeated. The robot detects the scene as a
revisited place and closes the loop, results in False Loop Closure. This
false loop closure destroys the robot's trajactory completely. Our Approach
based on odometric assistance will allow the robot to counter check the
position before closing the loop, results in avoiding the false loop closure in
an adversarial environment of repeated scene.
