English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

Monocular simultaneous localization and mapping (SLAM) is emerging inadvanced driver assistance systems and autonomous driving, because a singlecamera is cheap and easy to install. Conventional monocular SLAM has two majorchallenges leading inaccurate localization and mapping. First, it ischallenging to estimate scales in localization and mapping. Second,conventional monocular SLAM uses inappropriate mapping factors such as dynamicobjects and low-parallax areas in mapping. This paper proposes an improvedreal-time monocular SLAM that resolves the aforementioned challenges byefficiently using deep learning-based semantic segmentation. To achieve thereal-time execution of the proposed method, we apply semantic segmentation onlyto downsampled keyframes in parallel with mapping processes. In addition, theproposed method corrects scales of camera poses and three-dimensional (3D)points, using estimated ground plane from road-labeled 3D points and the realcamera height. The proposed method also removes inappropriate corner featureslabeled as moving objects and low parallax areas. Experiments with eight videosequences demonstrate that the proposed monocular SLAM system achievessignificantly improved and comparable trajectory tracking accuracy, compared toexisting state-of-the-art monocular and stereo SLAM systems, respectively. Theproposed system can achieve real-time tracking on a standard CPU potentiallywith a standard GPU support, whereas existing segmentation-aided monocular SLAMdoes not.

Improved Real-Time Monocular SLAM Using Semantic Segmentation on Selective Frames