UAV Obstacle Detection Algorithm Based on Improved ORB Sparse Optical Flow

Li, Pengyue; Hao, Xiangyang; Wang, Junqiang; Gu, Youyi; Wang, Gaojie

doi:10.1109/iaeac47372.2019.8997799

Cited by 1 publication

(3 citation statements)

References 1 publication

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Figures 7 and 8 below illustrates the filte dynamic objects from the freiburg3_Walking_xyz highly dynamic scene sequence TUM dataset. One of the men in the picture is walking randomly [28][29][30]. Because Yolov4-Tiny is a lightweight network, part of the detection accuracy is sacrificed to improve the running speed of the algorithm.…”

Section: Lk Optical Flow Constraintmentioning

confidence: 99%

“…Figures 7 and 8 below illustrates the filtering of dynamic objects from the freiburg3_Walking_xyz highly dynamic scene sequence in the TUM dataset. One of the men in the picture is walking randomly [28][29][30]. Figure 7 is the result of traditional orb-slam2 without dynamic feature point filtering, and Figure 8 is the result of dynamic feature point filtering.…”

Section: Lk Optical Flow Constraintmentioning

confidence: 99%

“…In the experiment, the mainstream public dataset TUM RGB-D was used to evaluate the performance of the SLAM algorithm proposed in this paper. This dataset is a standard RGB-D dataset provided by the Computer Vision Class group of Technical University of Munich, Germany, and it has been used by many scholars in the SLAM research field to evaluate the performance of SLAM algorithms [30][31][32][33][34]. The data in the dataset are mainly divided into the low dynamic scene dataset fr3_sitting_xx and the high dynamic scene dataset fr3_walking_xx.…”

Section: Experimental Data Setsmentioning

confidence: 99%

See 2 more Smart Citations

RGB-D Visual SLAM Based on Yolov4-Tiny in Indoor Dynamic Environment

Chang

Sun³

et al. 2022

Micromachines

View full text Add to dashboard Cite

For a SLAM system operating in a dynamic indoor environment, its position estimation accuracy and visual odometer stability could be reduced because the system can be easily affected by moving obstacles. In this paper, a visual SLAM algorithm based on the Yolov4-Tiny network is proposed. Meanwhile, a dynamic feature point elimination strategy based on the traditional ORBSLAM is proposed. Besides this, to obtain semantic information, object detection is carried out when the feature points of the image are extracted. In addition, the epipolar geometry algorithm and the LK optical flow method are employed to detect dynamic objects. The dynamic feature points are removed in the tracking thread, and only the static feature points are used to estimate the position of the camera. The proposed method is evaluated on the TUM dataset. The experimental results show that, compared with ORB-SLAM2, our algorithm improves the camera position estimation accuracy by 93.35% in a highly dynamic environment. Additionally, the average time needed by our algorithm to process an image frame in the tracking thread is 21.49 ms, achieving real-time performance.

show abstract

Section: Lk Optical Flow Constraintmentioning

confidence: 99%

Section: Lk Optical Flow Constraintmentioning

confidence: 99%