Kudan Visual SLAM (KdVisual) update: Wheel odometry improvement
09.03.2021
Wheel odometry – a basic but powerful sensor fusion
Many use-cases using simultaneous localization and mapping (SLAM) leverage several sensors to make the localization system more robust and reliable. Wheel odometry is one of the commonly available sensors in wheeled robots and vehicles. It provides feedback on the movement of a robot by tracking the rotation of each of the wheels on the robot. Wheel odometry information is quite complementary to the camera information, so fusing the two data sources improves robustness and makes tracking loss more difficult.
As we continue to improve our algorithm, we have significantly updated how Kudan Visual SLAM uses the wheel odometry information.
Wheel odometry helps Visual SLAM especially in dark featureless environments as well as dynamic environments
While camera-only Visual SLAM works very well in most cases and has a lot of benefits such as lower cost and not requiring GPS or specific infrastructure, it is not perfect. For example, Visual SLAM systems can have difficulties localizing or tracking if the operating environment doesn’t have a lot of features (such as rooms with plain white walls) or if it is in a simple dark environment, or if the scenery changes wildly from moment to moment. However, since wheel odometry only utilizes wheel rotation, it is insulated from the effects described above and can complement the vision system across varying conditions.
The following video demonstrates the effectiveness of fusing wheel odometry when there are intermittent gaps in the visual information and the ability of the SLAM system to maintain accurate position information.
Demo video
Kudan Visual SLAM: Wheel odometry improvement
We are continuing to improve accuracy and robustness so that our customers expand the use of KdVisual across multiple environments and sensor configurations reliably.
About Kudan Inc.
Kudan (Tokyo Stock Exchange securities code: 4425) is a Deep Tech research and development company specializing in algorithms to enable artificial perception (AP). As a complement to artificial intelligence (AI), AP functions allow machines to develop autonomy. Currently, Kudan is using its high-level technical innovation to explore business areas based on its milestone models established for Deep Tech, which provide a wide-ranging impact on several major industrial fields. For more information, please refer to Kudan’s website at https://www.kudan.io/.
■Company Details
Name: Kudan Inc.
Securities Code: 4425
Representative: CEO Daiu Ko
■For more details, please feel free to contact us from here.