ID 原文 译文
21135 最后，该文创新性地引入图像间显著性传播约束因子来克服超像素误匹配带来的影响。 Third, the inter-image saliency value propagation constraint parameter is innovatively introduced to overcome the disadvantages of superpixel mismatching.
21136 在公开测试数据集上的实验结果表明，所提算法在检测精度和检测效率上优于目前的主流算法，并具有较强的鲁棒性。 Experimental results on public test datasets show that the proposed algorithm issuperior over current state-of-the-art methods in terms of detection accuracy and detection efficiency, and hasstrong robustness.
21137 车辆检测是遥感图像分析领域的热点研究内容之一，车辆目标的智能提取和识别，对于交通管理、城市建设有重要意义。 Vehicle detection is one of the hotspots in the field of remote sensing image analysis. The intelligent extraction and identification of vehicles are of great significance to traffic management and urban construction.
21138 在遥感领域中，现有基于卷积神经网络的车辆检测方法存在实现过程复杂并且对于车辆密集区域检测效果不理想的缺陷。 In remote sensing field, the existing methods of vehicle detection based on Convolution Neural Network (CNN)are complicated and most of these methods have poor performance for dense areas.
21139 针对上述问题，该文提出基于端到端的神经网络模型DF-RCNN以提高车辆密集区域的检测精度。 To solve above problems, anend-to-end neural network model named DF-RCNN is presented to solve the detecting difficulty in dense areas.
21140 首先，在特征提取阶段，DF-RCNN模型将深浅层特征图的分辨率统一并融合； Firstly, the model unifies the resolution of the deep and shallow feature maps and combines them.
21141 其次，DF-RCNN模型结合可变形卷积和可变形感兴趣区池化模块，通过加入少量的参数和计算量以学习目标的几何形变。 After that,the deformable convolution and RoI pooling are used to study the geometrical deformation of the target byadding a small number of parameters and calculations.
21142 实验结果表明，该文提出的模型针对密集区域的车辆目标具有较好的检测性能。 Experimental results show that the proposed model hasgood detection performance for vehicle targets in dense areas.
21143 在当前的网络体系结构下，采用硬件系统实现服务器集群负载均衡存在着获取负载节点状态困难、流量导向方式复杂等制约因素，不利于提升服务器集群的伸缩性和服务性能。 Under the present network architecture, it is disadvantageous for scalability and service performanceof server cluster to adopt hardware systems to realize load balancing of server cluster, because there are somerestriction factors in such a method, including the difficulty of acquiring load nodes status and the complexityof redirecting traffic, etc.
21144 针对此问题，该文提出一种基于软件定义网络(SDN)的负载均衡机制(SDNLB)。 To solve the problem, a Load Balancing mechanism based on Software-DefinedNetworking (SDNLB) is proposed.