Geo-location for Ground Target with Multiple Observations Using Unmanned Aerial Vehicle

In order to improve the target location accuracy of unmanned aerial vehicle(UAV),a novel target location method using multiple observations is proposed.Firstly,the camera intrinsic parameters are calibrated.Then,the weighted least squares estimation is used to improve the localization precision because the traditional crossover method is vulnerable to noise and has low precision.By repeatedly measuring the same target point,a nonlinear observation equation is established and then covered to linear equations using Taylor expansion.The weighted matrix is obtained according to the height of the measurement point and the camera optic axis pointing angle,and then the weighted least squares estimation is used to calculate the target position iteratively.Finally,the effectiveness and robustness of this method is verified by numerical simulation and flight test.The results show that this method can effectively improve the precision of target location.

A NEW METHOD FOR DETERMINING THE LOCATION AND LENGTH OF A THIN CYLINDRICAL TARGET WITH RANGE DIFFERENCE OF A MULTISTATIC RADAR SYSTEM

This paper describes a short range target location system based on the range difference information of a T-R4 multistatic radar system with FMCW signal. A new method is proposed to determine the location and length of a thin cylindrical target making use of the high resolution of wide band FMCW signal and the spectrum characteristics of the target echo. Formulae are derived for target location and its length estimation being independent of the transmitter position. System performances are simulated with the proposed algorithm and the results are given for various situations.

Self-sustained target waves in excitable media with only a long-range link

In this paper we investigate spatiotemporal pattern formation in excitable media with only a long-range link. Besides the trivial solutions of spiral patterns, we find the asymptotic self-sustained target waves in the autonomous tissues. The wave source supporting this kind of new pattern is the oscillatory one-dimensional Winfree-loop self- organized under the presence of a long-range link, which is explored by the dominant phase-advanced driving method. Based on this understanding we can effectively regulate the oscillations of excitable media by suitably arranging the long-range link, including construction of self-sustained target waves with controllable period and wave length, or manipulation of system states between different patterns.

Scale effect removal and range migration correction for hypersonic target coherent detection

The detection of hypersonic targets usually confronts range migration(RM)issue before coherent integration(CI).The traditional methods aiming at correcting RM to obtain CI mainly considers the narrow-band radar condition.However,with the increasing requirement of far-range detection,the time bandwidth product,which is corresponding to radar’s mean power,should be promoted in actual application.Thus,the echo signal generates the scale effect(SE)at large time bandwidth product situation,influencing the intra and inter pulse integration performance.To eliminate SE and correct RM,this paper proposes an effective algorithm,i.e.,scaled location rotation transform(ScLRT).The ScLRT can remove SE to obtain the matching pulse compression(PC)as well as correct RM to complete CI via the location rotation transform,being implemented by seeking the actual rotation angle.Compared to the traditional coherent detection algorithms,Sc LRT can address the SE problem to achieve better detection/estimation capabilities.At last,this paper gives several simulations to assess the viability of ScLRT.

Subpixel Target Location Techniques for 3-D Coordinate Measuring System

the close photogrammetric 3-D coordinate measurement is a newmeasuring technology in the fields of the coordinate measurementmachine (CMM) in recent years. In this method, we usually place sometargets on the measured object and take image of targets to determinethe object coordinate. The subpixel location of target image plays animportant role in high accuracy 3-D coordinate measuring procedure.In this paper, some subpixel location methods are reviewed and somefactors which affect location precision are analyzed.

Combined treatment promotes the long-range axon regeneration to right brain targets

Axons in the peripheral nervous system(PNS)can regenerate after injury.However,the adult mammalian central nervous system(CNS)loses the intrinsic regrowth ability.No robust axon regeneration occurs spontaneously after nerve injury,which was clearly observed by Ramon y Cajal in the early 20^(th) century(1,2).Due to lack

A Method to Obtain the Moving Speed of Uncooperative Target Based on Only Two Measurements

The existing research results show that a fixed single station must conduct three consecutive frequency shift measurements and obtain the target’s moving speed by constructing two frequency difference equations. This article proposes a new method that requires only two consecutive measurements. While using the azimuth measurement to obtain the angular difference between two radial distances, it also conducts two consecutive Doppler frequency shift measurements at the same target azimuth. On the basis of this measurement, a frequency difference equation is first constructed and solved jointly with the Doppler frequency shift equation. By eliminating the velocity variable and using the measured angular difference to obtain the target’s lead angle, the target’s velocity can be solved by using the Doppler frequency shift equation again. The new method avoids the condition that the target must move equidistantly, which not only provides an achievable method for engineering applications but also lays a good foundation for further exploring the use of steady-state signals to achieve passive positioning.

Vehicle Real-time Location Based on Visual Perception Model

Vehicle recognition system (VRS) plays a very important role in the field of intelligent transportation systems.A novel and intuitive method is proposed for vehicle location.The method we provide for vehicle location is based on human visual perception model technique. The perception color space HSI in this algorithm is adopted.Three color components of a color image and more potential edge patterns are integrated for solving the feature extraction problem.A fast and automatic threshold technique based on human visual perception model is also developed.The vertical edge projection and horizontal edge projection are adopted for locating left-right boundary of vehicle and top-bottom boundary of vehicle, respectively. Very promising experimental results are obtained using real-time vehicle image sequences, which have confirmed that this proposed location vehicle method is efficient and reliable, and its calculation speed meets the needs of the VRS.

基于改进YOLO v5的复杂环境下桑树枝干识别定位方法

为实现复杂自然环境下对桑树嫩叶处枝干的识别检测,改变当前桑叶采摘设备作业过程中依赖人工辅助定位的现状,解决识别目标姿态多样和环境复杂导致的低识别率问题,提出一种基于改进YOLO v5模型的桑树枝干识别模型(YOLO v5-mulberry),并结合深度相机构建定位系统。首先,在YOLO v5的骨干网络中加入CBAM(Convolutional block attention module)注意力机制,提高神经网络对桑树枝干的关注度;并增加小目标层使模型可检测4像素×4像素的目标,提高了模型检测小目标的性能;同时使用GIoU损失函数替换原始网络中的IoU损失函数,有效防止了预测框和真实框尺寸较小时无法正确反映预测框及真实框之间位置关系的情况;随后,完成深度图和彩色图的像素对齐,通过坐标系转换获取桑树枝干三维坐标。试验结果表明:YOLO v5-mulberry检测模型的平均精度均值为94.2%,较原模型提高16.9个百分点,置信度也提高12.1%;模型室外检测时应检测目标数53,实际检测目标数为48,检测率为90.57%;桑树嫩叶处枝干三维坐标识别定位系统的定位误差为(9.4985 mm,11.285 mm,19.11 mm),满足使用要求。该研究可实现桑树嫩叶处枝干的识别与定位,有助于推动桑叶智能化采摘机器人研究。

基于COF-YOLOv5s的油茶果识别定位

针对自然环境下油茶果存在多种类、多遮挡及小目标等问题,该研究基于YOLOv5s提出COFYOLOv5s(camellia oleifera fruit-you only look once)油茶果识别模型,实现油茶果的高精度检测。通过添加小目标检测层、将FasterNet中的轻量化模块Faster Block嵌入到C3模块及添加Biformer注意力机制对YOLOv5s进行改进。试验结果表明,改进后网络在测试集上的精度、召回率、平均精度均值分别为97.6%、97.8%、99.1%,比YOLOv5s分别提高5.0、7.5、4.4个百分点,推理时间为10.3 ms。将模型部署到Jetson Xavier NX中,结合ZED mini相机进行油茶果识别与定位试验。室内试验得到COF-YOLOv5s的召回率为91.7%,室外绿油茶果的召回率为68.8%,小目标红油茶果在弱光条件下的召回率为64.3%。研究结果可为实现油茶产业的智能化和规模化提供理论支持。

基于顶点与主体区域同步检测的精准车牌定位

为应对非约束环境下的车牌精定位问题,提出一种基于顶点局部区域与主体区域同步检测策略的非约束性车牌定位算法。通过删减YOLOv5网络的输出结构,训练得到可同步检测车牌及顶点区域的车牌检测网络,在兼顾精度与计算速度的前提下,实现车牌顶点和主体区域的同步定位。针对一幅图中存在多个车牌区域及顶点区域存在少量漏检和误检的情况,分别设计了车牌顶点归类和单一缺失顶点预测后处理算法,借助顶点间的空间位置关系进行漏检目标预测和误检目标排查,有效改善了因场景复杂导致的个别顶点目标检测效果差的问题。所提算法在中国城市停车场数据集(CCPD)上的测试结果显示,平均精准率达99.25%,平均召回率达98.70%。所提算法不仅能够准确预测出车牌的4个顶点坐标,而且在中端GPU硬件平台上处理速度可达121帧/s,具有较好的应用价值。

基于神经网络的工业机器人视觉抓取系统设计

针对机器人示教编程方法导致的工件位置固定、抓取效率低下的问题,研究神经网络在机器人视觉识别与抓取规划中的应用,建立了视觉引导方案,通过YOLOV5神经网络模型开发视觉识别系统,识别物体的种类,同时获取待抓取物体定位点坐标;提出了机器人六点手眼标定原理并进行标定实验,提出了针对俯视图为圆形或长方形物体的定位方法;最后针对3种物体进行了180次的抓取实验,实验的综合平均抓取成功率约为92.8%,验证了视觉识别和抓取机器人系统具备实际应用的可能性,有效提高了抓取效率。

异型圆柱贯穿腔体多轴线定位方法研究

异型圆柱贯穿腔体承载件的多轴线定位是其自动化缺陷检测的前提,本文提出一种基于机器视觉的异型圆柱贯穿腔体承载件的多轴线定位方法,从根源上解决了人工定位测量速度慢、精度低、自动化定位不能批量处理多个特征等问题。首先标定相机内参和机械臂手眼关系,接着采集异型承载件图像并检测端面椭圆,然后识别端面ArUco靶标并估计相机位姿,最后将承载件端面圆心坐标和法向量转换到机器人基座坐标系下,求解轴线数学表征。实验结果表明,该方法可以快速定位承载件轴线特征,异型承载件端面圆心定位精度最高可达1.63mm,端面ArUco靶标姿态估计精度最高可达0.35°,效率远超于人工测量,并可以处理具有多个圆柱端面的异型工件,具有一定的工业应用价值。

海面目标光电单站无源定位与误差分析

为实现光电侦察设备对海面舰船目标单站无源定位,提出了对海面目标的高精度近实时光电单站无源定位方法。建立了单站无源定位方程,分析了目标定位误差,探讨了影响定位误差因素,给出了单站无源定位的应用建议。仿真结果表明:在17.899 km距离处单站无源定位的定位误差约为1.672%R,单次定位时间约20 ms;位置、纬度、经度和高度误差分别服从N(299.31 m,197.39 m)、N(4.944×10^(-5°),2.665×10^(-3°))、N(4.038×10^(-5°),1.899×10^(-3°))、N(-0.485 m,14.165 m)的正态分布;俯仰角、方位角和目标高度的误差对定位精度影响较大,俯仰角在[-120°,-15°]范围内定位结果可靠,目标高度误差在[-200 m,200 m]范围内,定位误差小于800 m,实现了对海面目标的高精度近实时单站无源定位。提出的方法适用于具备高度参考面估计的海面、地面和空中目标的单站无源定位。

非合作线谱声源分布式定位方法

非合作目标定位是水声定位领域的研究热点。为充分挖掘并利用各类可观测参数,提高非合作目标定位能力,提出一种基于目标频率变化信息的非合作目标定位方法。该方法针对匀速直线运动目标的定位问题,首先根据多普勒频移原理,建立观测频率与目标辐射频率、目标运动速度、位置之间的函数映射关系,利用最小均方准则建立目标函数,通过优化算法估计分布式定位系统中各个测量单元与目标运动轨迹的致近点距离,最后综合各观测节点的测距结果,构建几何定位模型,求得目标运动轨迹的解析解。通过仿真分析,证明了该方法的有效性,并指出了该方法需要目标通过与各测量单元的致近点才能获得较好的距离估计能力。文中分析了不同频率估计精度对定位精度的影响,结果表明,提出的定位方法对测频精度具有一定的容限,是一种高精度的非合作线谱声源定位方法。

隐伏振动源的地震定位技术

隐伏目标探测是军事防护的重点之一,但是难以借助常规的光学或图像技术实现。针对这一问题,本文首先回溯和总结地震定位技术的发展和应用,展示了地震方法在隐伏振动目标探测和定位上具有显著的优势;然后利用中国柴达木盆地西部采集的微地震资料,采用非线性定位法和双差法对微地震事件进行了定位。最终结果表明:1)与非线性定位方法相比,重定位的微震震源分布得更集中,震群之间的边界更清晰,与已知的断裂构造相符;2)重定位的误差降低了一个量级,绝大多数地震事件经过非线性定位后的水平和深度误差在1000 m以内,而双差法的误差则控制在100 m以内。该实例证实了利用地震定位技术探测隐伏振动目标的可行性。

自然资源视频监测目标检测关键技术研究与应用

随着机器视觉技术的突飞猛进,智能视频监测为自然资源监管提供了可实时响应的新执法手段。运用基于深度学习的域自适应目标识别技术和基于DEM辅助的高精度单目定位技术,对自然资源违法行为与线索进行快速目标检测;并在此基础上设计了自然资源视频执法监测系统技术架构,在广东开展了实地应用测试。结果表明,该方法对重型机械、土堆、建筑板材、在建砖房、新建板房、红砖堆等类别的识别准确率均高于80%,在1 200 m监测距离内平均定位误差小于3%。

一种融合时差频差和测向的运动目标跟踪方法

传统的星载无源定位系统对空中辐射源定位求解通常采用假设高程的方法,高程假设误差将对定位跟踪精度造成较大影响。为实现未知高程运动辐射源的高精度定位跟踪,针对异轨三星构型的无源定位系统,提出了一种基于时差、频差和二维测向融合的迭代扩展卡尔曼滤波(Iterative Extended Kalman Filter,IEKF)跟踪方法。在WGS-84坐标系下建立了状态方程和观测方程,并采用IEKF方法对目标状态进行估计。仿真结果表明,该方法可对未知高程的运动目标进行高精度状态估计,典型仿真场景下的目标高程估计精度达到百米量级,相对于已有方法收敛时间更短,并且在卫星覆盖范围内具有更大的高精度定位跟踪区域。

视觉引导机器人的电池镍片识别与标定方法研究

针对电池PACK线电池镍片摆放难以自动化问题,以某电池自动化生产线中机器人对电池镍片的识别与无序抓取为例,进行了理论分析和测试试验研究。用Eye-to-Hand的三点手眼标定法建立了机器人与视觉相机之间的仿射变换矩阵;利用Halcon软件平台对拍摄电池镍片图像进行零件特征提取和基于形状的模板匹配算法进行镍片的识别,并研究了镍片各种位姿情况下对识别速度、准确性的影响;最后搭建了测试平台进行了验证。研究结果表明,该方法的识别速度达到了32.3 ms,识别准确度达到了0.9521,一次性可以识别10个以上不同位姿形态的镍片,可以精准有效地引导机器人对物料进行操作,满足了该自动化生产线的需求,对类似作业类型有较好的参考价值。

基于对称性评价的曲线对称中心拟合算法

在缺乏形状先验信息的前提下,对有噪声干扰的一维图像(灰度曲线)进行中心位置估计是机器视觉测量中的关键问题之一。以曲线与其镜像的匹配误差作为对称性评价函数,提出一种对称中心拟合算法,并采用最小二乘法计算最佳匹配点作为对称中心。该算法在使用中需要迭代计算,并可能收敛至错误的位置,通过对迭代初始值选取可能产生的局部收敛进行分析,证明了局部极值点仅可能出现在真值相邻的半像素之内,就此提出了收敛点验证策略,解决了错误收敛的问题。通过仿真模拟与实例,验证了对称中心拟合算法在各种干扰下的稳健性,在强噪声干扰下,灰度质心法的位置变化量达到了数十甚至上百像素,而对称中心拟合算法的位置变化量均方根值仍能保持在1像素左右,在其他噪声条件下,对称中心拟合算法的表现也远优于灰度质心法。