This paper focuses on the effective utilization of data augmentation techniques for 3Dlidar point clouds to enhance the performance of neural network models.These point clouds,which represent spatial information throu...This paper focuses on the effective utilization of data augmentation techniques for 3Dlidar point clouds to enhance the performance of neural network models.These point clouds,which represent spatial information through a collection of 3D coordinates,have found wide-ranging applications.Data augmentation has emerged as a potent solution to the challenges posed by limited labeled data and the need to enhance model generalization capabilities.Much of the existing research is devoted to crafting novel data augmentation methods specifically for 3D lidar point clouds.However,there has been a lack of focus on making the most of the numerous existing augmentation techniques.Addressing this deficiency,this research investigates the possibility of combining two fundamental data augmentation strategies.The paper introduces PolarMix andMix3D,two commonly employed augmentation techniques,and presents a new approach,named RandomFusion.Instead of using a fixed or predetermined combination of augmentation methods,RandomFusion randomly chooses one method from a pool of options for each instance or sample.This innovative data augmentation technique randomly augments each point in the point cloud with either PolarMix or Mix3D.The crux of this strategy is the random choice between PolarMix and Mix3Dfor the augmentation of each point within the point cloud data set.The results of the experiments conducted validate the efficacy of the RandomFusion strategy in enhancing the performance of neural network models for 3D lidar point cloud semantic segmentation tasks.This is achieved without compromising computational efficiency.By examining the potential of merging different augmentation techniques,the research contributes significantly to a more comprehensive understanding of how to utilize existing augmentation methods for 3D lidar point clouds.RandomFusion data augmentation technique offers a simple yet effective method to leverage the diversity of augmentation techniques and boost the robustness of models.The insights gained from this research can pave the way for future work aimed at developing more advanced and efficient data augmentation strategies for 3D lidar point cloud analysis.展开更多
Ventilation system analysis for underground mines has remained mostly unchanged since the Atkinson method was made popular by Mc Elroy in 1935. Data available to ventilation technicians and engineers is typically limi...Ventilation system analysis for underground mines has remained mostly unchanged since the Atkinson method was made popular by Mc Elroy in 1935. Data available to ventilation technicians and engineers is typically limited to the quantity of air moving through any given heading. Because computer-aided modelling, simulation, and ventilation system design tools have improved, it is now important to ensure that developed models have the most accurate information possible. This paper presents a new technique for estimating underground drift friction factors that works by processing 3 D point cloud data obtained by using a mobile Li DAR. Presented are field results that compare the proposed approach with previously published algorithms, as well as with manually acquired measurements.展开更多
Nowadays,the deep learning for object detection has become more popular and is widely adopted in many fields.This paper focuses on the research of LiDAR and camera sensor fusion technology for vehicle detection to ens...Nowadays,the deep learning for object detection has become more popular and is widely adopted in many fields.This paper focuses on the research of LiDAR and camera sensor fusion technology for vehicle detection to ensure extremely high detection accuracy.The proposed network architecture takes full advantage of the deep information of both the LiDAR point cloud and RGB image in object detection.First,the LiDAR point cloud and RGB image are fed into the system.Then a high-resolution feature map is used to generate a reliable 3D object proposal for both the LiDAR point cloud and RGB image.Finally,3D box regression is performed to predict the extent and orientation of vehicles in 3D space.Experiments on the challenging KITTI benchmark show that the proposed approach obtains ideal detection results and the detection time of each frame is about 0.12 s.This approach could establish a basis for further research in autonomous vehicles.展开更多
基金funded in part by the Key Project of Nature Science Research for Universities of Anhui Province of China(No.2022AH051720)in part by the Science and Technology Development Fund,Macao SAR(Grant Nos.0093/2022/A2,0076/2022/A2 and 0008/2022/AGJ)in part by the China University Industry-University-Research Collaborative Innovation Fund(No.2021FNA04017).
文摘This paper focuses on the effective utilization of data augmentation techniques for 3Dlidar point clouds to enhance the performance of neural network models.These point clouds,which represent spatial information through a collection of 3D coordinates,have found wide-ranging applications.Data augmentation has emerged as a potent solution to the challenges posed by limited labeled data and the need to enhance model generalization capabilities.Much of the existing research is devoted to crafting novel data augmentation methods specifically for 3D lidar point clouds.However,there has been a lack of focus on making the most of the numerous existing augmentation techniques.Addressing this deficiency,this research investigates the possibility of combining two fundamental data augmentation strategies.The paper introduces PolarMix andMix3D,two commonly employed augmentation techniques,and presents a new approach,named RandomFusion.Instead of using a fixed or predetermined combination of augmentation methods,RandomFusion randomly chooses one method from a pool of options for each instance or sample.This innovative data augmentation technique randomly augments each point in the point cloud with either PolarMix or Mix3D.The crux of this strategy is the random choice between PolarMix and Mix3Dfor the augmentation of each point within the point cloud data set.The results of the experiments conducted validate the efficacy of the RandomFusion strategy in enhancing the performance of neural network models for 3D lidar point cloud semantic segmentation tasks.This is achieved without compromising computational efficiency.By examining the potential of merging different augmentation techniques,the research contributes significantly to a more comprehensive understanding of how to utilize existing augmentation methods for 3D lidar point clouds.RandomFusion data augmentation technique offers a simple yet effective method to leverage the diversity of augmentation techniques and boost the robustness of models.The insights gained from this research can pave the way for future work aimed at developing more advanced and efficient data augmentation strategies for 3D lidar point cloud analysis.
基金supported by the Natural Sciences and Engineering Research Council of Canada (NSERC) under grant CRDPJ 44580412Barrick Gold Corporation and Peck Tech Consulting Ltd
文摘Ventilation system analysis for underground mines has remained mostly unchanged since the Atkinson method was made popular by Mc Elroy in 1935. Data available to ventilation technicians and engineers is typically limited to the quantity of air moving through any given heading. Because computer-aided modelling, simulation, and ventilation system design tools have improved, it is now important to ensure that developed models have the most accurate information possible. This paper presents a new technique for estimating underground drift friction factors that works by processing 3 D point cloud data obtained by using a mobile Li DAR. Presented are field results that compare the proposed approach with previously published algorithms, as well as with manually acquired measurements.
基金This work was supported by the National Key Research and Development Program of China(2017YFB0102603,2018YFB0105003)the National Natural Science Foundation of China(51875255,61601203,61773184,U1564201,U1664258,U1764257,U1762264)+3 种基金the Natural Science Foundation of Jiangsu Province(BK20180100)the Six Talent Peaks Project of Jiangsu Province(2018-TD-GDZB-022)the Key Project for the Development of Strategic Emerging Industries of Jiangsu Province(2016-1094)the Key Research and Development Program of Zhenjiang City(GY2017006).
文摘Nowadays,the deep learning for object detection has become more popular and is widely adopted in many fields.This paper focuses on the research of LiDAR and camera sensor fusion technology for vehicle detection to ensure extremely high detection accuracy.The proposed network architecture takes full advantage of the deep information of both the LiDAR point cloud and RGB image in object detection.First,the LiDAR point cloud and RGB image are fed into the system.Then a high-resolution feature map is used to generate a reliable 3D object proposal for both the LiDAR point cloud and RGB image.Finally,3D box regression is performed to predict the extent and orientation of vehicles in 3D space.Experiments on the challenging KITTI benchmark show that the proposed approach obtains ideal detection results and the detection time of each frame is about 0.12 s.This approach could establish a basis for further research in autonomous vehicles.
