| ID | 原文 | 译文 |
| 7114 | 最后利用相关数据,对所提出的方法进行应用研究,验证了方法的有效性。 | Finally, using the related data, research and application of the proposed method to verify the effectiveness of the method. |
| 7115 | 基于二项分布的扩展目标概率假设密度(extended target probability hypothesis density based on binominal distribution,BET-PHD)算法能够获得比泊松ET-PHD更好的跟踪性能。 | Extended target based on the binomial distribution probability hypothesis density (extended target aim-listed probability content density -based on binominal distribution, BET - PHD) algorithm can obtain a better tracking performance than poisson ET - PHD. |
| 7116 | 然而,BET-PHD中作为先验信息的检测概率和量测数目最大值在实际应用中是未知的。 | BET - PHD as a priori information, however, the maximum number of detection probability and the measurement is unknown in the practical application. |
| 7117 | 参数严重不匹配会导致算法性能急剧下降。 | Parameter severe mismatch leads to algorithm performance fell sharply. |
| 7118 | 鉴于已有文献给出量测数目最大值的估计方法,提出一种能够在线估计检测概率的贝塔高斯ET-PHD(beta Gaussian ET-PHD,BG-ET-PHD)滤波器。 | Given the existing literature method, the maximum number of estimation is given in this paper, a to estimate detection probability of Gaussian ET beta - online PHD (beta Gaussian ET - PHD, BG - ET - PHD) filter. |
| 7119 | 首先采用二项分布的共轭先验贝塔分布估计检测概率,并与BET-PHD相结合得到BG-ET-PHD。 | First USES the binomial distribution of the conjugate prior distribution of beta estimate detection probability, and combined with a BET - PHD get BG - ET - PHD. |
| 7120 | 仿真结果表明,BG-ET-PHD滤波器能够准确估计检测概率,能够获得比基于泊松模型的伽马高斯ET-PHD(gamma Gaussian ET-PHD,GG-ET-PHD)更好的跟踪性能。 | The simulation results show that BG - ET - PHD filter can accurately estimate the probability of detection, to obtain than gamma Gaussian based on poisson model ET - PHD (gamma Gaussian ET - PHD, GG - ET - PHD) better tracking performance. |
| 7121 | 实现对目标的自动检测与跟踪是坦克火控系统未来发展的重要方向。 | Realize the goal of automatic detection and tracking is the important direction of the future development of tank fire control system. |
| 7122 | 首先采用迁移学习的方法将基于深度学习模型的Faster R-卷积神经网络(faster R-convolution neural network,Faster R-CNN)算法应用解决复杂背景下的坦克装甲目标检测问题,与基于人工模型的传统算法相比达到了较高的检测精度。 | First adopt the method of migration study will be based on the model of deep learning Faster - R - convolution neural network (Faster - R - convolution neural network, the Faster - R - CNN) algorithm solve the problem of tank armored targets detection under complicated background, compared with the traditional algorithm based on artificial model reached high precision. |
| 7123 | 其次,针对坦克火控系统现有目标跟踪算法的不足,通过将Faster R-CNN算法与现有跟踪算法相结合,提出了复合式目标跟踪算法,实现了对坦克装甲目标的自动检测与稳定跟踪。 | Secondly, aiming at the shortcomings of the existing target tracking algorithm of tank fire control system, by using a Faster - R - CNN algorithm combined with the existing tracking algorithm, multiple target tracking algorithm is proposed and realized with the automatic detection and stability of tank armor target tracking. |