SORB: improve ORB feature matching by semantic segmentation

Hao Chen; Kaiwei Wang; Weijian Hu; Lei Fei

doi:10.1117/12.2325423

Translator Disclaimer

4 October 2018 SORB: improve ORB feature matching by semantic segmentation

Hao Chen; Kaiwei Wang; Weijian Hu; Lei Fei

Author Affiliations +

Proceedings Volume 10799, Emerging Imaging and Sensing Technologies for Security and Defence III; and Unmanned Sensors, Systems, and Countermeasures; 107990Z (2018) https://doi.org/10.1117/12.2325423
Event: SPIE Security + Defence, 2018, Berlin, Germany

Abstract

Feature matching is at the base of many computer vision algorithms such as SLAM, which is a technology widely used in the area from intelligent vehicles (IV) to assistance for the visually impaired (VI). This article presents an improved detector and a novel semantic-visual descriptor, coined SORB (Semantic ORB), combining binary semantic labels and traditional ORB descriptor. Compared to the original ORB feature, the new SORB performs better in uniformity of distribution and accuracy of matching. We demonstrate it through experiments on some open source datasets and several real-world images obtained by RealSense.

Citation Download Citation

Hao Chen, Kaiwei Wang, Weijian Hu, and Lei Fei "SORB: improve ORB feature matching by semantic segmentation", Proc. SPIE 10799, Emerging Imaging and Sensing Technologies for Security and Defence III; and Unmanned Sensors, Systems, and Countermeasures, 107990Z (4 October 2018); https://doi.org/10.1117/12.2325423