Akademik Veri Yönetim Sistemi
7th International Symposium on Multidisciplinary Studies and Innovative Technologies, ISMSIT 2023, Ankara, Türkiye, 26 - 28 Ekim 2023
This study focuses on the development and evaluation of a visual odometry system, which incorporates feature matching algorithms, such as scale-invariant feature transform (SIFT) and binary robust independent elementary features (BRIEF). The system aims to achieve precise localization in autonomous driving scenarios by accurately estimating camera motion and constructing reliable vehicle trajectories. Extensive comparative evaluations were conducted using datasets obtained from the onboard cameras of the CARLA simulator to assess the performance of the visual odometry system. The results highlight the effectiveness of feature matching techniques, in achieving high-quality and dependable localization. The trajectory estimation capability of the visual odometry system plays a crucial role in enhancing the safety and efficiency of autonomous vehicles. The integration of SIFT and BRIEF algorithms within the visual odometry pipeline demonstrates their significance in extracting and matching features for precise localization. The evaluation on CARLA datasets showcases the potential of the visual odometry system as a fundamental component in autonomous driving applications. Its successful implementation can greatly enhance the overall autonomous driving experience by providing reliable and precise localization. This research demonstrates how the visual odometry system, equipped with feature matching algorithms, trajectory estimation capabilities, and the utilization of SIFT and BRIEF, achieves accurate and reliable localization in autonomous driving scenarios using data from the CARLA simulator. The comparison results can viewed by clicking the youtube link below.1