Pub Date : 2022-12-05DOI: 10.1109/ROBIO55434.2022.10011979
Yao Yang, Jing Wang, Yi Zhang, Xiao-jing Liu
This paper investigates the relationship between the navigation accuracy and the distance from the tracing regions to the registration center. The investigation was carried out using a 3D-printed plastic model and an infrared binocular camera. 7 registration points were set up around the nasal-orbital area as registration points. 7 groups of checking points were used to test the registration accuracy, which was at a distance of 10, 20, 30, 40, 50, 60, and 70 mm from the registration area center, respectively. The accuracy of each group of checking points was investigated. SPSS 19.0 was used to calculate the mean error of each tracing point. The tracing error of different areas was compared by t-test. The relationship between distance and error was investigated by the linear regression method. P<0.05 was considered as significant difference. Results: The deviation of navigation points in the 7 groups of registered checking points is 0.737±0.236 mm, with the largest deviation at 1.307 mm, the smallest at 0.272 mm, and the mean 95% CI at (0.6296, 0.8449). The navigation error (y) and the distance from the registration center (x) coincided with the linear regression, the regression equation was identified to be y=-0.451+0.178x. This regression model is statistically significant (P<0.05). The navigation error increases when the tracing region moves far away from the registration area.
{"title":"Experimental Investigations on the Relationship Between the Navigation Accuracy and the Tracing Distance from the Registration Center","authors":"Yao Yang, Jing Wang, Yi Zhang, Xiao-jing Liu","doi":"10.1109/ROBIO55434.2022.10011979","DOIUrl":"https://doi.org/10.1109/ROBIO55434.2022.10011979","url":null,"abstract":"This paper investigates the relationship between the navigation accuracy and the distance from the tracing regions to the registration center. The investigation was carried out using a 3D-printed plastic model and an infrared binocular camera. 7 registration points were set up around the nasal-orbital area as registration points. 7 groups of checking points were used to test the registration accuracy, which was at a distance of 10, 20, 30, 40, 50, 60, and 70 mm from the registration area center, respectively. The accuracy of each group of checking points was investigated. SPSS 19.0 was used to calculate the mean error of each tracing point. The tracing error of different areas was compared by t-test. The relationship between distance and error was investigated by the linear regression method. P<0.05 was considered as significant difference. Results: The deviation of navigation points in the 7 groups of registered checking points is 0.737±0.236 mm, with the largest deviation at 1.307 mm, the smallest at 0.272 mm, and the mean 95% CI at (0.6296, 0.8449). The navigation error (y) and the distance from the registration center (x) coincided with the linear regression, the regression equation was identified to be y=-0.451+0.178x. This regression model is statistically significant (P<0.05). The navigation error increases when the tracing region moves far away from the registration area.","PeriodicalId":151112,"journal":{"name":"2022 IEEE International Conference on Robotics and Biomimetics (ROBIO)","volume":"71 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-12-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129208703","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-12-05DOI: 10.1109/ROBIO55434.2022.10011890
Dong Zhang, Yan Gai, Renjie Ju, Zhiwen Miao, Ju Lao
Cable-driven manipulators (CDMs) are widely used for operations in confined spaces due to their slender bodies and multiple degrees of freedom (DOFs). To plan passable paths for them in narrow spaces, a rapidly exploring random tree (RRT) algorithm is often used. However, the cost of planning process are not considered in this method. In order to improve the quality of path planning of CDMs, this work optimizes a traditional RRT algorithm by fusing it with an A* algorithm. In the novel RRT-A* method, the RRT algorithm is used to generate feasible paths, the A* algorithm is used to estimate the cost and measure the selection of the traversal search of each feasible node of the path. Compared with the traditional RRT algorithm, the novel algorithm is better in some performances such as complex path, path redundancy and large random path angle. Simulation results show that this method can effectively reduce path cost and the number of nodes. For further validation, a 17 DOFs CDM prototype is conducted to move in multi-obstacle environments to test the proposed method.
{"title":"A RRT-A* Path Planning Algorithm for Cable-driven Manipulators","authors":"Dong Zhang, Yan Gai, Renjie Ju, Zhiwen Miao, Ju Lao","doi":"10.1109/ROBIO55434.2022.10011890","DOIUrl":"https://doi.org/10.1109/ROBIO55434.2022.10011890","url":null,"abstract":"Cable-driven manipulators (CDMs) are widely used for operations in confined spaces due to their slender bodies and multiple degrees of freedom (DOFs). To plan passable paths for them in narrow spaces, a rapidly exploring random tree (RRT) algorithm is often used. However, the cost of planning process are not considered in this method. In order to improve the quality of path planning of CDMs, this work optimizes a traditional RRT algorithm by fusing it with an A* algorithm. In the novel RRT-A* method, the RRT algorithm is used to generate feasible paths, the A* algorithm is used to estimate the cost and measure the selection of the traversal search of each feasible node of the path. Compared with the traditional RRT algorithm, the novel algorithm is better in some performances such as complex path, path redundancy and large random path angle. Simulation results show that this method can effectively reduce path cost and the number of nodes. For further validation, a 17 DOFs CDM prototype is conducted to move in multi-obstacle environments to test the proposed method.","PeriodicalId":151112,"journal":{"name":"2022 IEEE International Conference on Robotics and Biomimetics (ROBIO)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-12-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129225544","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Multi-dimensional force sensors are of great significance to improve the perception of robots. It's very important to remove the drift and noise of the multi-dimensional force sensor signal caused by environmental changes. Recurrent Neural Network based on Long-Short Term Memory (LSTM-RNN) is proposed for real-time signal processing of multi-dimensional force sensors. Firstly, Adaptive Empirical Mode Decomposition (AEMD) is verified to be effective in removing drift and noise from multi-dimensional force sensor signals. Then, AEMD is utilized to process the force sensor signal and LSTM-RNN is trained by the processed signal. In the force test experiment, the errors of different signals processed by LSTM-RNN are very small and smaller than those of RNN signal processing, which proves that the trained LSTM-RNN can effectively process multi-dimensional force sensor signals in real time.
{"title":"Real-time processing of force sensor signals based on LSTM-RNN","authors":"Qiao Liu, Yu Dai, Mengwen Li, Bin Yao, Yunwei Xin, Jianxun Zhang","doi":"10.1109/ROBIO55434.2022.10011703","DOIUrl":"https://doi.org/10.1109/ROBIO55434.2022.10011703","url":null,"abstract":"Multi-dimensional force sensors are of great significance to improve the perception of robots. It's very important to remove the drift and noise of the multi-dimensional force sensor signal caused by environmental changes. Recurrent Neural Network based on Long-Short Term Memory (LSTM-RNN) is proposed for real-time signal processing of multi-dimensional force sensors. Firstly, Adaptive Empirical Mode Decomposition (AEMD) is verified to be effective in removing drift and noise from multi-dimensional force sensor signals. Then, AEMD is utilized to process the force sensor signal and LSTM-RNN is trained by the processed signal. In the force test experiment, the errors of different signals processed by LSTM-RNN are very small and smaller than those of RNN signal processing, which proves that the trained LSTM-RNN can effectively process multi-dimensional force sensor signals in real time.","PeriodicalId":151112,"journal":{"name":"2022 IEEE International Conference on Robotics and Biomimetics (ROBIO)","volume":"89 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-12-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125404170","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
With the popularity and intelligence of electric vehicle, the increasing demand for charging convenience has driven the development of automatic charging technology. The recognition and localization of electric vehicle charging socket is the key to automatic charging. This study proposes a system for fast recognition and localization of electric vehicle charging socket based on deep learning and affine correction. First, modify the yolov4 network structure for recognizing the charging socket to improve the recognition speed. Second, using the meanshift clustering algorithm, the noise is effectively removed to improve the recognition success rate. Third, we propose a pixel coordinate correction method for the charging socket based on the affine transformation. The projective transformation is approximated to the affine transformation when the camera is facing the charging socket. According to the properties of covariance and distance ratio invariance, the pixel coordinates of the charging holes are corrected. Finally, the charging socket is located by the Perspective-n-Point (PnP) algorithm. With different angles, distances and light intensities, the recognition success rate of the charging socket is 100%, and the average recognition time for single-frame image is 27ms. The localization accuracy is tested under different light intensity and distances. After affine correction, the localization accuracy is improved, and the final average localization errors are 1.418 degrees, 1.660 degrees, 0.050 degrees, 0.217mm, 0.215mm and 0.855mm in Rx, Ry, Rz, x, y and $z$ respectively. The results show that our method has a good effect on the recognition and localization of the charging socket in complex environment.
{"title":"Fast Recognition and Localization of Electric Vehicle Charging Socket Based on Deep Learning and Affine Correction","authors":"Peiyuan Zhao, Xiaopeng Chen, Shengquan Tang, Yang Xu, Mingming Yu, Peng Xu","doi":"10.1109/ROBIO55434.2022.10011985","DOIUrl":"https://doi.org/10.1109/ROBIO55434.2022.10011985","url":null,"abstract":"With the popularity and intelligence of electric vehicle, the increasing demand for charging convenience has driven the development of automatic charging technology. The recognition and localization of electric vehicle charging socket is the key to automatic charging. This study proposes a system for fast recognition and localization of electric vehicle charging socket based on deep learning and affine correction. First, modify the yolov4 network structure for recognizing the charging socket to improve the recognition speed. Second, using the meanshift clustering algorithm, the noise is effectively removed to improve the recognition success rate. Third, we propose a pixel coordinate correction method for the charging socket based on the affine transformation. The projective transformation is approximated to the affine transformation when the camera is facing the charging socket. According to the properties of covariance and distance ratio invariance, the pixel coordinates of the charging holes are corrected. Finally, the charging socket is located by the Perspective-n-Point (PnP) algorithm. With different angles, distances and light intensities, the recognition success rate of the charging socket is 100%, and the average recognition time for single-frame image is 27ms. The localization accuracy is tested under different light intensity and distances. After affine correction, the localization accuracy is improved, and the final average localization errors are 1.418 degrees, 1.660 degrees, 0.050 degrees, 0.217mm, 0.215mm and 0.855mm in Rx, Ry, Rz, x, y and $z$ respectively. The results show that our method has a good effect on the recognition and localization of the charging socket in complex environment.","PeriodicalId":151112,"journal":{"name":"2022 IEEE International Conference on Robotics and Biomimetics (ROBIO)","volume":"70 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-12-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126241565","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-12-05DOI: 10.1109/ROBIO55434.2022.10011835
F. Asano, Haosong Chen, Runyu Liu
This paper proposes a method for achieving ultrahigh-speed strict stealth walking (USSW) of a planar combined rimless wheel (CRW) with 2-DOF wobbling mass. In the first half, a stable USSW gait generation for the CRW on a non-slip road surface is investigated. We develop a 7-DOF mathematical model, and designing a strict output following control so that the entire COM position moves in the walking direction at a constant speed and the next stance foot can land on the ground stealthily. The numerical simulation shows that the resultant force of the horizontal ground reaction forces becomes zero according to the method. In the latter half, we introduce another model that added a rod to synchronize the rotational motion of the fore and rear legs with the aim of achieving USSW on the road surface where the coefficient of friction is zero. It is numerically shown that a stable USSW gait can be generated according to the modified output following control, but there is a problem that the vertical ground reaction force becomes negative during motion when the walking speed is very high.
{"title":"Ultrahigh-speed Strict Stealth Walking of Combined Rimless Wheel with 2-DOF Wobbling Mass","authors":"F. Asano, Haosong Chen, Runyu Liu","doi":"10.1109/ROBIO55434.2022.10011835","DOIUrl":"https://doi.org/10.1109/ROBIO55434.2022.10011835","url":null,"abstract":"This paper proposes a method for achieving ultrahigh-speed strict stealth walking (USSW) of a planar combined rimless wheel (CRW) with 2-DOF wobbling mass. In the first half, a stable USSW gait generation for the CRW on a non-slip road surface is investigated. We develop a 7-DOF mathematical model, and designing a strict output following control so that the entire COM position moves in the walking direction at a constant speed and the next stance foot can land on the ground stealthily. The numerical simulation shows that the resultant force of the horizontal ground reaction forces becomes zero according to the method. In the latter half, we introduce another model that added a rod to synchronize the rotational motion of the fore and rear legs with the aim of achieving USSW on the road surface where the coefficient of friction is zero. It is numerically shown that a stable USSW gait can be generated according to the modified output following control, but there is a problem that the vertical ground reaction force becomes negative during motion when the walking speed is very high.","PeriodicalId":151112,"journal":{"name":"2022 IEEE International Conference on Robotics and Biomimetics (ROBIO)","volume":"16 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-12-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121565629","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Joint calibration is one of the fundamental works to ensure the locomotion performance of quadrupedal robots. Inaccurate joint offset calibration accuracy will incur foot-tip position errors and significant disturbances to locomotion performance, especially in highly dynamic scenarios. This paper proposes an accurate joint offset calibration method for quadrupedal robots. We derive the leg kinematic error model based on the product of the exponentials formula and use the iterative least squares algorithm to obtain the joint offset of the quadrupedal robot. Considering the influence of the body frame on the calibration of the abduction/adduction (Ab/Ad) joint, the offset of the Ab/Ad joint is modified by the angle between the z-axis of the body frame and that of the leg frame. We verify the effectiveness of the proposed method on an experimental quadrupedal robot, where the maximum foot-tip position error is decreased from 13.97mm to 2.25mm after calibration.
{"title":"Accurate Joint Offset Calibration for Quadrupedal Robots","authors":"Chuanlin Zhao, Letian Qian, Shuhan Wang, Qi Li, Huaxing Wang, Xinhao Luo","doi":"10.1109/ROBIO55434.2022.10011907","DOIUrl":"https://doi.org/10.1109/ROBIO55434.2022.10011907","url":null,"abstract":"Joint calibration is one of the fundamental works to ensure the locomotion performance of quadrupedal robots. Inaccurate joint offset calibration accuracy will incur foot-tip position errors and significant disturbances to locomotion performance, especially in highly dynamic scenarios. This paper proposes an accurate joint offset calibration method for quadrupedal robots. We derive the leg kinematic error model based on the product of the exponentials formula and use the iterative least squares algorithm to obtain the joint offset of the quadrupedal robot. Considering the influence of the body frame on the calibration of the abduction/adduction (Ab/Ad) joint, the offset of the Ab/Ad joint is modified by the angle between the z-axis of the body frame and that of the leg frame. We verify the effectiveness of the proposed method on an experimental quadrupedal robot, where the maximum foot-tip position error is decreased from 13.97mm to 2.25mm after calibration.","PeriodicalId":151112,"journal":{"name":"2022 IEEE International Conference on Robotics and Biomimetics (ROBIO)","volume":"34 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-12-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122996822","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-12-05DOI: 10.1109/ROBIO55434.2022.10011964
Jie Yang, Ziqi Zhao, Xuesu Xiao, Jiankun Wang, M. Meng
Early detection of adolescent idiopathic scoliosis (AIS) is essential for AIS treatment and prevention of AIS progression. However, the existing clinical scoliosis assessment method, the standing full-column radiographs (X-ray) imaging, is radioactive, making this method unsuitable for large-scale promotion among adolescents. As a result, many countries have implemented school scoliosis screening programs (SSS) to achieve large-scale scoliosis screening and monitoring of adolescents by measuring the angle of trunk rotation (ATR). However, the SSS is time-consuming and inaccurate due to subjective manual examination. In this paper, we present an automatic method to calculate ATR based on the contour curve of the human back. This automatic method begins with a 3D depth sensor-scanned point cloud model of the human back and identifies the spinous process and stress points by obtaining the back contour curve from the depth information. Finally, the ATR is calculated according to the measurement principle of scoliosis meter. We demonstrate the effectiveness of our method using twenty-seven pairs of ATR data from nine participants with AFBT. There is not only a significant positive correlation, but also a convinced level of agreement between ATRs obtained using automatic method and ATRs obtained using manual method in the SSS. The experiment results reveal that the proposed method can efficiently achieve accurate measurement of ATR in the SSS.
{"title":"Automatic Angle of Trunk Rotation Detection Using 3D Sensor Imaging in Scoliosis Assessment","authors":"Jie Yang, Ziqi Zhao, Xuesu Xiao, Jiankun Wang, M. Meng","doi":"10.1109/ROBIO55434.2022.10011964","DOIUrl":"https://doi.org/10.1109/ROBIO55434.2022.10011964","url":null,"abstract":"Early detection of adolescent idiopathic scoliosis (AIS) is essential for AIS treatment and prevention of AIS progression. However, the existing clinical scoliosis assessment method, the standing full-column radiographs (X-ray) imaging, is radioactive, making this method unsuitable for large-scale promotion among adolescents. As a result, many countries have implemented school scoliosis screening programs (SSS) to achieve large-scale scoliosis screening and monitoring of adolescents by measuring the angle of trunk rotation (ATR). However, the SSS is time-consuming and inaccurate due to subjective manual examination. In this paper, we present an automatic method to calculate ATR based on the contour curve of the human back. This automatic method begins with a 3D depth sensor-scanned point cloud model of the human back and identifies the spinous process and stress points by obtaining the back contour curve from the depth information. Finally, the ATR is calculated according to the measurement principle of scoliosis meter. We demonstrate the effectiveness of our method using twenty-seven pairs of ATR data from nine participants with AFBT. There is not only a significant positive correlation, but also a convinced level of agreement between ATRs obtained using automatic method and ATRs obtained using manual method in the SSS. The experiment results reveal that the proposed method can efficiently achieve accurate measurement of ATR in the SSS.","PeriodicalId":151112,"journal":{"name":"2022 IEEE International Conference on Robotics and Biomimetics (ROBIO)","volume":"191 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-12-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123006582","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-12-05DOI: 10.1109/ROBIO55434.2022.10011936
Julian Bialas, M. Döller
Coverage path planning (CPP) for unmanned aerial vehicles (UAVs) defines a vital role in the automation process of UAV-supported disaster management. While multiple algorithms exist to solve the CPP problem for planar areas, the proposed algorithm is the first to handle complex three-dimensional environments and also account for power constraints and changing environments. By applying proximal policy optimization to an advantage-based actor-critic deep reinforcement learning model, the proposed framework enables an agent to efficiently cover the target area (TA), considering the orientation of the observation sensor, avoiding collisions as well as no-flying zones (NFZ) and reacting to changing environments. Furthermore, a safe landing mechanism, based on the Dijkstra algorithm, expands the framework to guarantee a successful landing in the respective start and landing zone (SLZ) within the power constraints. The model is trained on real data to learn the optimal control policy. Additionally, the framework was tested and validated on hardware in a drone lab to confirm its effectiveness and capability to perform real-time path planning.
{"title":"Coverage Path Planning for Unmanned Aerial Vehicles in Complex 3D Environments with Deep Reinforcement Learning","authors":"Julian Bialas, M. Döller","doi":"10.1109/ROBIO55434.2022.10011936","DOIUrl":"https://doi.org/10.1109/ROBIO55434.2022.10011936","url":null,"abstract":"Coverage path planning (CPP) for unmanned aerial vehicles (UAVs) defines a vital role in the automation process of UAV-supported disaster management. While multiple algorithms exist to solve the CPP problem for planar areas, the proposed algorithm is the first to handle complex three-dimensional environments and also account for power constraints and changing environments. By applying proximal policy optimization to an advantage-based actor-critic deep reinforcement learning model, the proposed framework enables an agent to efficiently cover the target area (TA), considering the orientation of the observation sensor, avoiding collisions as well as no-flying zones (NFZ) and reacting to changing environments. Furthermore, a safe landing mechanism, based on the Dijkstra algorithm, expands the framework to guarantee a successful landing in the respective start and landing zone (SLZ) within the power constraints. The model is trained on real data to learn the optimal control policy. Additionally, the framework was tested and validated on hardware in a drone lab to confirm its effectiveness and capability to perform real-time path planning.","PeriodicalId":151112,"journal":{"name":"2022 IEEE International Conference on Robotics and Biomimetics (ROBIO)","volume":"107 ","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-12-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"120970327","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-12-05DOI: 10.1109/ROBIO55434.2022.10011785
Chaohong He, Yang Gao, Xingben Wang
In order to achieve high-precision positioning of unmanned vehicles in low-light environments, based on the system framework of the VINS-Fusion algorithm, a fusion positioning algorithm LL- VI G for unmanned vehicles under low-light conditions is proposed. Aiming at the problems of low contrast, noise, and difficulty in feature extraction under low-light conditions, A multi-layer fusion image enhancement algorithm is proposed to improve the number of corner points extracted under low light conditions. For the problems of cumulative error in VI-SLAM and GNSS signals being easily interfered, a graph optimization method is used to integrate the GNSS global image. The fusion of positioning information and VI-SLAM positioning results reduces the cumulative error of VI-SLAM to a certain extent, and at the same time provides high-precision positioning in the absence of GNSS signals, improving the positioning accuracy and robustness of unmanned vehicles. The multi-layer fusion image enhancement algorithm proposed in this paper is experimentally verified based on the New Tsukuba Stereo dataset. The results show that the image enhanced by this algorithm can effectively increase the number of corner extractions. The LL-VIG algorithm proposed in this paper is experimentally verified based on the KITTI public data set and real vehicle scenarios. The results show that the positioning accuracy of LL- VI G is significantly higher than that of the comparison algorithm VINS-Fusion.
为了实现低光环境下无人车的高精度定位，在VINS-Fusion算法的系统框架基础上，提出了一种低光条件下无人车的融合定位算法LL- VI G。针对低光照条件下图像对比度低、噪声大、特征提取困难等问题，提出了一种多层融合图像增强算法，提高了低光照条件下提取的角点数量。针对VI-SLAM图像累积误差大、GNSS信号易受干扰的问题，采用图优化方法对GNSS全局图像进行整合。定位信息与VI-SLAM定位结果的融合在一定程度上减小了VI-SLAM的累积误差，同时在没有GNSS信号的情况下提供高精度定位，提高了无人车的定位精度和鲁棒性。本文提出的多层融合图像增强算法在新筑波立体数据集上进行了实验验证。结果表明，该算法增强后的图像可以有效地增加角点提取的次数。基于KITTI公共数据集和真实车辆场景，对本文提出的LL-VIG算法进行了实验验证。结果表明，LL- VI G的定位精度明显高于比较算法VINS-Fusion。
{"title":"Research on Fusion Localization Algorithm of Unmanned Vehicles under Low Light Conditions","authors":"Chaohong He, Yang Gao, Xingben Wang","doi":"10.1109/ROBIO55434.2022.10011785","DOIUrl":"https://doi.org/10.1109/ROBIO55434.2022.10011785","url":null,"abstract":"In order to achieve high-precision positioning of unmanned vehicles in low-light environments, based on the system framework of the VINS-Fusion algorithm, a fusion positioning algorithm LL- VI G for unmanned vehicles under low-light conditions is proposed. Aiming at the problems of low contrast, noise, and difficulty in feature extraction under low-light conditions, A multi-layer fusion image enhancement algorithm is proposed to improve the number of corner points extracted under low light conditions. For the problems of cumulative error in VI-SLAM and GNSS signals being easily interfered, a graph optimization method is used to integrate the GNSS global image. The fusion of positioning information and VI-SLAM positioning results reduces the cumulative error of VI-SLAM to a certain extent, and at the same time provides high-precision positioning in the absence of GNSS signals, improving the positioning accuracy and robustness of unmanned vehicles. The multi-layer fusion image enhancement algorithm proposed in this paper is experimentally verified based on the New Tsukuba Stereo dataset. The results show that the image enhanced by this algorithm can effectively increase the number of corner extractions. The LL-VIG algorithm proposed in this paper is experimentally verified based on the KITTI public data set and real vehicle scenarios. The results show that the positioning accuracy of LL- VI G is significantly higher than that of the comparison algorithm VINS-Fusion.","PeriodicalId":151112,"journal":{"name":"2022 IEEE International Conference on Robotics and Biomimetics (ROBIO)","volume":" 9","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-12-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"120971453","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-12-05DOI: 10.1109/ROBIO55434.2022.10011697
Yukun Liu, Ruiqing Luo, Minghui He, Liang Du, Sheng Bao, Jianjun Yuan, Weiwei Wan
The study of friction has long been popular with scientists. Considering the coupling characteristic of multi-input-multi-output (MIMO) system, the coupling friction model was established, which was based on Coulomb-Viscous friction model for the differential modular robot joint (DMRJ) according to the law of conservation of energy. Then, we identified the coefficients of the friction model through the experiment. In order to verify the accuracy of the established friction model, we regarded the DMRJ as a 2-DoF linkage and built the inertial dynamic model based on Lie theory, whose inertial parameters were estimated by computer aided design (CAD). Besides, we chose the trajectory based on the Fourier series, which has good performance in anti-interference ability, as the verification trajectory. From the generated trajectories, we used a trajectory that is able to change as much as possible in speed, position and acceleration to estimate the accuracy of the model comprehensively. Finally, the result indicated the model had good accuracy.
{"title":"Dynamic modeling and analysis for a differential modular robot joint with the friction model","authors":"Yukun Liu, Ruiqing Luo, Minghui He, Liang Du, Sheng Bao, Jianjun Yuan, Weiwei Wan","doi":"10.1109/ROBIO55434.2022.10011697","DOIUrl":"https://doi.org/10.1109/ROBIO55434.2022.10011697","url":null,"abstract":"The study of friction has long been popular with scientists. Considering the coupling characteristic of multi-input-multi-output (MIMO) system, the coupling friction model was established, which was based on Coulomb-Viscous friction model for the differential modular robot joint (DMRJ) according to the law of conservation of energy. Then, we identified the coefficients of the friction model through the experiment. In order to verify the accuracy of the established friction model, we regarded the DMRJ as a 2-DoF linkage and built the inertial dynamic model based on Lie theory, whose inertial parameters were estimated by computer aided design (CAD). Besides, we chose the trajectory based on the Fourier series, which has good performance in anti-interference ability, as the verification trajectory. From the generated trajectories, we used a trajectory that is able to change as much as possible in speed, position and acceleration to estimate the accuracy of the model comprehensively. Finally, the result indicated the model had good accuracy.","PeriodicalId":151112,"journal":{"name":"2022 IEEE International Conference on Robotics and Biomimetics (ROBIO)","volume":"47 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-12-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121706525","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：