Work

CUDA-based SeqSLAM for Real-Time Place Recognition

Year: 2017
Type: conference-paper
Abstract: Vehicle localization is a fundamental issue in autonomous navigation that has been extensively studied by the\nRobotics community. An important paradigm for vehicle localization is based on visual place recognition which\nrelies on learning a database, then consecutively trying to find matchings between this database and the actual\nvisual input. An increasing interest has been directed to visual place recognition in varying conditions like day\nand night cycles and seasonal changes. A major approach dealing with such challenges is Sequence SLAM (SeqSLAM)\nbased on matching a sequence of images to the database instead of a single image. This algorithm allows\nglobal pose recovery at the expense of a higher computational time. To solve this problem with a certain amount\nof speedup, we propose in this work, a CUDA-based solution for real-time place recognition with SeqSLAM.\nWe design a mapping of SeqSLAM to CUDA architecture and we describe, in detail, our hardware-specific implementation\nconsiderations as well as the parallelization methods. Performance analysis against existing CPU\nimplementation is also given, showing a speedup to six times faster than the CPU for common sized databases.\nMore speedup could be obtained when dealing with bigger databases.
Source: HAL
Language: English
Open Access status: green
OpenAlex ID: W3193335385
FWCI: 0.068
Cites: 21
Cited by: 3
Related to: 10