Twinny Robot

Blog

Twinny’s indoor autonomous cart - Core Technology of Autonomous Cart - Object Tracking Using Depth Camera

JANUARY 11, 2019

  1. Object tracking through depth camera is necessary for Twinny’s indoor autonomous cart to recognize and follow the user.


  2. Twinny’s indoor autonomous cart


Object Tracking Using Depth Camera

Our superior method of object tracking enables autonomous cart to consistently track registered objects with little computational load.

This technology uses depth cameras to accurately track pre-registered objects in complex environments. 

This technology, to recognize specific objects in complex environments, is essential, but challenging research.



For example, if there are bunch of objects which have similar appearance with its target, or if there are illumination changes, it gets confused and ends up losing the target.
To overcome these problems, we pre-process the image so that it is robust to environmental changes, and collectively use the information of the object- such as the size, shape, and motion - to track the target.
As a result, Twinny’s autonomous cart distinguishes multiple objects and pinpoints the target using 3D environmental data and color images from the depth camera.