空拍機的感測器融合與互補式自我定位

Sensor Fusion and Complementary Ego-Positioning of Drone

沈天翼(Tian-Yi Shen)

指導教授 ： 洪一平

國立臺灣大學/電機資訊學院/資訊網路與多媒體研究所/碩士(2016年)

https://doi.org/10.6342/NTU201701866

摘要

即時定位與地圖構建(SLAM)是一種解決飛行器自我定位問題的通用方法。但是，目前現有的SLAM方法，都有各種各樣的限制，比如在缺少特徵點，有重複圖案的場景中，或是由於相機的動態模糊的影響，都會使得SLAM定位結果缺失或是偏移。因此，爲了剋服僅基於視覺的飛行器自我定位方法的局限，我們提出了一種互補式的定位方法。該方法基於拓展卡爾曼濾波，將IMU作爲額外的傳感器融合進定位過程中，並基於IMU的讀數，來修正視覺定位的結果。同時，該方法將基於特徵點以及直接法這兩種不同分類的SLAM定位方法進行融合，進行互補式定位。實驗表明，我們的方法不僅可以提升視覺定位的結果，同時在視覺定位出現大的失誤時，可以轉換成另一個進行互補式定位，使飛行器的自我定位更穩定。

關鍵字

互補式定位 ； 即時定位與地圖構建 ； 飛行器

並列摘要

Visual simultaneous localization and mapping (SLAM) is a common solution for ego-positioning of drone. However, SLAM may sometimes lose tracking due to fast camera motions, featureless or repetitive environments, etc. To overcome the limitations of vision-only SLAM, we propose a new complementary method in this paper, which fuses the visual positioning results with those estimated by an inertial measurement sensor (IMU) based on a loosely-coupled framework, and further combines feature-based SLAM with direct SLAM by underperformance detection to keep advantages of both methods, i.e. not only keeping the performance of feature-based SLAM but also overcoming featureless or motion-blurred issues with direct SLAM. Experiments on both simulated and real datasets show that the proposed method improves the accuracy of current vision-only SLAM and also leads to more robust positioning results with complementary framework.

並列關鍵字

Visual inertial odometry ； Complementary ； Drone

參考文獻

[2] Georg Klein and David Murray, “Parallel tracking and mapping for small AR workspaces,” in Mixed and Augmented Reality, 2007. ISMAR 2007.6th IEEE and ACM International Symposium on. IEEE, pp. 225–234, 2007.

[4] Richard A Newcombe, Steven J Lovegrove, and Andrew J Davison, “Dtam: Dense tracking and mapping inreal-time,” in 2011 international conference on computer vision. IEEE, pp.2320–2327, 2011.

[6] Raul Mur-Artal, JMM Montiel, and Juan D Tardós, “ORB SLAM: a versatile and accurate monocular SLAM system,” IEEE Transactions on Robotics, vol.31, no.5, pp. 1147–1163, 2015.

[7] Kuan-Wen Chen, Chun-Hsin Wang, Xiao Wei, Qiao Liang, Chu-Song Chen, Ming-Hsuan Yang, and Yi Ping Hung, “Vision-based positioning for internet-of vehicles,” IEEE Transactions on Intelligent Transportation Systems, 2016.

[9] Jonathan Kelly and Gaurav S Sukhatme, “Visual inertial simultaneous localization, mapping and sensor-to-sensor self-calibration,” in Computational Intelligence in Robotics and Automation (CIRA), 2009 IEEE International Symposium on. IEEE, pp.360–368, 2009.