Paper Title :“Visual Odometry by Intel Real Sense T265 by 3 Different Views”

Author :Shivani Baldwa, Raghav Jethliya, Chuang Jan Chang

Article Citation :Shivani Baldwa ,Raghav Jethliya ,Chuang Jan Chang , (2020 ) " “Visual Odometry by Intel Real Sense T265 by 3 Different Views” " , International Journal of Electrical, Electronics and Data Communication (IJEEDC) , pp. 59-62, Volume-8,Issue-9

Abstract : MOIL-SDK can reference different spatial angles to distill multiple-images conforming to those of central perspective from a fisheye image. Accordingly, we extend an open monocular visual odometry project to generate 3 odometry that tracks from a fisheye video. It is a real-time, precise, and persistent approach. The contributions of our work are to create a MOIL video dataset using Intel RealSense T265 camera as it associated with an embedded IMU sensor that produces 6DOF that helps to generate ground truth data for the camera moving. We also modify the regular single-tracked video odometer to multiple tracks. With this extension, it is forecast able that the inherent error accumulation issue consults to a single trajectory will have a chance to improve. Keywords - MOIL-SDK, Intel RealSense T265, Visual Odometry: Fisheye image, Camera Calibration.

Type : Research paper

Published : Volume-8,Issue-9

	|		PDF	|	Viewed - 69	|	Published on 2021-01-07