In this paper, the author says that during SLAM process, pseudo segments that appear from any momentary pause of dynamic objects in laser data would make the map unsatisfied.

How is this caused?

If the dynamic object moved, won't laser data update and eliminate the segment of dynamic objects?


I think you're misunderstanding the following two sentences from the abstract:

(i) the first one is a fusion module which synthesizes line segments obtained from laser rangefinder and line features extracted from monocular camera. This policy eliminates any pseudo segments that appear from any momentary pause of dynamic objects in laser data.

It looks like they're first extracting lines from the point cloud generated from the laser range finder. They then cross-reference these lines with lines extracted via computer vision on the monocular image. If a line extracted from the point cloud doesn't exist in the image, then it is NOT a valid line. Therefore, you are correct, "If the dynamic object moved, [the] laser data [WILL] update and eliminate the segment" caused by dynamic objects. However, they're trying to correctly identify segments specifically when the dynamic object is briefly static.

You can see a specific example on page 441 where a person standing still caused a segment to be extracted from the point cloud. Fortunately, this segment does not exist in the monocular image, and so it was deemed an INVALID segment. There's also a brief description of this process in section 4 on page 437.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.