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
Three-dimensional(3D)lidar has been widely used in various fields.The MEMS scanning system is one of its most important components,while the limitation of scanning angle is the main obstacle to improve the demerit for...Three-dimensional(3D)lidar has been widely used in various fields.The MEMS scanning system is one of its most important components,while the limitation of scanning angle is the main obstacle to improve the demerit for its application in various fields.In this paper,a folded large field of view scanning optical system is proposed.The structure and parameters of the system are determined by theoretical derivation of ray tracing.The optical design software Zemax is used to design the system.After optimization,the final structure performs well in collimation and beam expansion.The results show that the scan angle can be expanded from±5°to±26.5°,and finally the parallel light scanning is realized.The spot diagram at a distance of 100 mm from the exit surface shows that the maximum radius of the spot is 0.506 mm with a uniformly distributed spot.The maximum radius of the spot at 100 m is 19 cm,and the diffusion angle is less than 2 mrad.The energy concentration in the spot range is greater than 90%with a high system energy concentration,and the parallelism is good.This design overcomes the shortcoming of the small mechanical scanning angle of the MEMS lidar,and has good performance in collimation and beam expansion.It provides a design method for large-scale application of MEMS lidar.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the Shenzhen Fundamental Research Program(Grant No.JCYJ2020109150808037)the National Key Scientific Instrument and Equipment Development Projects of China(Grant No.62027823)the National Natural Science Foundation of China(Grant No.61775048)。
文摘Three-dimensional(3D)lidar has been widely used in various fields.The MEMS scanning system is one of its most important components,while the limitation of scanning angle is the main obstacle to improve the demerit for its application in various fields.In this paper,a folded large field of view scanning optical system is proposed.The structure and parameters of the system are determined by theoretical derivation of ray tracing.The optical design software Zemax is used to design the system.After optimization,the final structure performs well in collimation and beam expansion.The results show that the scan angle can be expanded from±5°to±26.5°,and finally the parallel light scanning is realized.The spot diagram at a distance of 100 mm from the exit surface shows that the maximum radius of the spot is 0.506 mm with a uniformly distributed spot.The maximum radius of the spot at 100 m is 19 cm,and the diffusion angle is less than 2 mrad.The energy concentration in the spot range is greater than 90%with a high system energy concentration,and the parallelism is good.This design overcomes the shortcoming of the small mechanical scanning angle of the MEMS lidar,and has good performance in collimation and beam expansion.It provides a design method for large-scale application of MEMS lidar.
