采样频率和激光脉宽对全波形激光雷达测距精度的影响

全波形激光雷达测距精度,又称测距重复精度或测距标准差,受激光器出光稳定性、激光脉宽、探测器响应时间抖动、电路噪声、波形形态、波形采样频率和波形处理算法等因素影响。理论分析了不同采样频率和不同脉宽对全波形激光雷达测距精度的影响,并采集不同的采样频率(1.25、2.5、5 GHz)和不同脉宽(1、2、3、···、10 ns)条件下的波形数据,经滤波、插值、波形提取等预处理后,利用线性高斯拟合、加权线性高斯拟合、迭代加权线性高斯拟合、期望最大化算法、和Levenberg Marquardt算法共5种算法计算测距值并统计测距精度。实验结果表明,EM算法获得的测距精度相比其他4种算法受到波形畸变的影响最小;加权线性高斯拟合算法获得的测距精度受采样频率变化的影响最小;相同波形幅值条件下,实际脉宽增加2.47倍,利用EM算法获得的测距精度从0.97 mm下降至1.18 mm,因此增加脉宽会降低测距精度;在光脉宽为4 ns的情况下,5 GHz采样频率数据在EM算法获得的测距精度分别为2.5 GHz、1.25 GHz采样频率数据的测距精度的1.71倍和3.07倍,而当2.5 GHz和1.25 GHz采样频率数据分别插值2倍和4倍至5 GHz后,仅为1.17倍和1.29倍,因此提高采样频率能够提高测距精度,而对低采样频率数据进行插值能够获得接近高采样频率数据的测距精度。     

Efficient graph attentional network for 3D object detection from Frustum-based LiDAR point clouds

LiDAR-based 3D object detection is important for autonomous driving scene perception, but point clouds produced by LiDAR are irregular and unstructured in nature, and cannot be adopted by the conventional Convolutional Neural Networks (CNN). Recently, Graph Convolutional Networks (GCN) has been proved as an ideal way to handle non-Euclidean structure data, as well as for point cloud processing. However, GCN involves massive computation for searching adjacent nodes, and the heavy computational cost limits its applications in processing large-scale LiDAR point cloud in autonomous driving. In this work, we adopt a frustum-based point cloud-image fusion scheme to reduce the amount of LiDAR point clouds, thus making the GCN-based large-scale LiDAR point clouds feature learning feasible. On this basis, we propose an efficient graph attentional network to accomplish the goal of 3D object detection in autonomous driving, which can learn features from raw LiDAR point cloud directly without any conversions. We evaluate the model on the public KITTI benchmark dataset, the 3D detection mAP is 63.72% on KITTI Cars, Pedestrian and Cyclists, and the inference speed achieves 7.9 fps on a single GPU, which is faster than other methods of the same type.     

CCNet: CNN model with channel attention and convolutional pooling mechanism for spatial image steganalysis

Image steganalysis based on convolutional neural networks(CNN) has attracted great attention. However, existing networks lack attention to regional features with complex texture, which makes the ability of discrimination learning miss in network. In this paper, we described a new CNN designed to focus on useful features and improve detection accuracy for spatial-domain steganalysis. The proposed model consists of three modules: noise extraction module, noise analysis module and classification module. A channel attention mechanism is used in the noise extraction module and analysis module, which is realized by embedding the SE(Squeeze-and-Excitation) module into the residual block. Then, we use convolutional pooling instead of average pooling to aggregate features. The experimental results show that detection accuracy of the proposed model is significantly better than those of the existing models such as SRNet, Zhu-Net and GBRAS-Net. Compared with these models, our model has better generalization ability, which is critical for practical application.     

Relational attention-based Markov logic network for visual navigation

The Journal of Supercomputing - We argue the agent’s low generalization problem for searching target object in challenging visual navigation could be solved by "how" and...     

Syntax–Aware graph convolutional network for the recognition of chinese implicit inter-sentence relations

The Journal of Supercomputing - In the literature, most previous studies on English implicit inter-sentence relation recognition only focused on semantic interactions, which could not exploit the...     

一种面向低采样率的点云数据处理网络

在直接处理点云的三维神经网络中,采样阶段实现了对原始点云中关键点的筛选,对于整个网络的性能及网络的抗噪能力具有重要作用。目前主流的最远点采样（FPS）方法在处理大规模3D点云数据时计算量大且耗时,并且低采样率时经过FPS采样后模型性能下降明显。针对这两个问题,提出一种面向低采样率的点云数据处理网络AS-Net。设计一个新的采样模块代替原backbone中的FPS,其由两个Layer组成,每个Layer基于长短期记忆网络获取原始点云与采样点云之间的联系权重,从而高效提取关键信息,去除冗余信息。在此基础上,利用注意力机制选择特征值较高的原始点云作为采样点,采样点作为后序任务的关键点输入到网络,进一步提高网络模型性能。基于ModelNet40数据集的实验结果表明,在低采样率条件下,AS-Net仍可达到81.6%的分类准确率,与使用FPS作为采样方法的网络模型相比提高52.7%。此外,其对噪声干扰具有很强的鲁棒性,对于大场景的分割时间效率优于同类采样方法。     

A prediction approach of SLM based on the ensemble of metamodels considering material efficiency,energy consumption,and tensile strength

Journal of Intelligent Manufacturing - As a rapid developing additive manufacturing (AM) technology, selective laser melting (SLM) provides a promising way for intelligent manufacturing. The SLM...     

scenario modeling for government big data governance decision-making: Chinese experience with public safety services

In the public safety service context, government big data governance (GBDG) is a challenging decision-making problem that encompasses uncertainties in the arenas of big data and its complex links. Modeling and collaborating the key scenario information required for GBDG decision-making can minimize system uncertainties. However, existing scenario-building methods are limited by their rigidity as they are employed in various application contexts and the associated high costs of modeling. In this paper, using a design science paradigm, a model-driven scenario modeling approach is proposed to achieve flexible scenario modeling for various applications through the transfer of generic domain knowledge. The key component of the proposed approach is a scenario meta-model that is built from existing literatures and practices by integrating qualitative, quantitative, and meta-modeling analysis. An instantiation mechanism of the scenario meta-model is also proposed to generate customized scenarios under Antecedent-Behavior-Consequence (ABC) theory. Two real-world safety service cases in Wuhan, China were evaluated to find that the proposed approach reduces GBDG decision-making uncertainties significantly by providing key information for GBDG problem identification, solution design, and solution value perception. This scenario-building approach can be further used to develop other GBDG systems for public safety services with reduced uncertainties and complete decision-making functions.     

A tight coupling mapping method to integrate the ESKF,g2o,and point cloud alignment

The Journal of Supercomputing - High-precision point cloud maps have drawn increasing attention due to their wide range of applications. In recent decades, point cloud maps are normally generated...     

Occluded Video Instance Segmentation: A Benchmark

International Journal of Computer Vision - Can our video understanding systems perceive objects when a heavy occlusion exists in a scene? To answer this question, we collect a large-scale dataset...