ID 原文 译文
10284 提取实时图中的MSER特征,与上下文地标基于规则化互相关准则进行特征匹配,利用双层匹配矫正策略减少误匹配,得到匹配特征对。 Landmark MSER of real-time image feature extracting, with the context based on the regulation rule of cross-correlation characteristics matching, using the double match correct strategy to reduce false matching, the matched features.
10285 提取匹配特征对的中心点作为参考点求解基准图与实时图之间的空间映射关系,进而利用最小二乘拟合一次多项式计算实时图中目标的位置坐标。 As a reference point to solve the center of the extraction and matching characteristics of the space mapping relationship between the reference image and real-time image, and then using the least squares fitting polynomial of calculating the location of the object in real-time image coordinates.
10286 实验结果表明,针对复杂地面场景,该方法的最大相对定误差不大于3个像素。 The experimental results show that, in view of the complex scene on the ground, the method of maximum relative error is not more than three pixels.
10287 基本满足成像末制导对自动目标识别算法稳健性好、识别精度高、抗干扰能力强等要求。 Basically met imaging terminal-guidance on automatic target recognition algorithm good robustness, high accuracy, strong anti-jamming capability, etc.
10288 目标运动状态的改变将导致目标跟踪算法精度降低或发散。 Target motion state changes will lead to a target tracking algorithm is reduced or divergent.
10289 为了提高机动目标跟踪的跟踪性能,首先,针对当前统计(current statistical,CS)模型中最大加速度固定设置导致模型误差增大的问题,提出了一种自适应CS模型; ‭In order to improve the tracking performance of maneuvering target tracking, first of all, in view of the current statistics, the current statistical, CS) model to the maximum acceleration of fixed setting model error increases in problem, an adaptive CS model is put forward;
10290 在自适应CS模型和交互式多模型(interacting multiple model,IMM)的基础上,提出了一种交互式多自适应模型(interacting multiple adaptive model,IMAM)。 In adaptive CS model and interactive multiple model (interacting multiple model, the IMM), on the basis of an interactive multiple adaptive model (interacting multiple adaptive model, IMAM).
10291 该模型通过采用两个自适应CS模型,能够有效消除目标状态突变造成模型误差急速增大的问题,提高了模型的准确度和适应性。 The model by adopting two adaptive CS model, can effectively eliminate the target state mutations cause the problem of model error increases rapidly, improve the accuracy and adaptability of the model.
10292 其次,在IMAM的基础上,结合修正卡尔曼滤波(amendatory Kalman filter,AKF)的思想,提出了IMAM-AKF算法。 ‭Secondly, on the basis of IMAM, combining with the modified Kalman filter (amendatory Kalman filter, AKF) thoughts, IMAM - AKF algorithm is proposed.
10293 该算法通过修正最终的状态融合估计值,有效地降低了目标机动造成的模型误差,进一步提高了机动目标跟踪的性能。 The algorithm through the fixed end state fusion estimation, effectively reduce the target maneuvering model error caused by the further improve the performance of the maneuvering target tracking.