基于目标跟踪的红外触摸屏优化算法

针对现有红外触摸屏多点识别时识别错乱和画图时轨迹毛刺严重等问题,在原有识别算法的基础上,加入了卡尔曼滤波与跟踪门算法结合的目标跟踪算法处理,提出一种新型红外触摸屏优化算法。该算法通过数据关联中的跟踪门逻辑,将量测值与已有的目标轨迹进行逻辑判断,选择正确的触点信息,剔除伪触点的信息,实现轨迹关联和目标跟踪。同时在红外触摸屏上建立触点的运动模型上,通过卡尔曼滤波实现轨迹的平滑和一定程度上的轨迹预测。在红外触摸屏上的测试表明,与原有的识别算法相比,优化算法增加了约3 μs单点识别时间,但是提高了轨迹在拐角处的平滑度,减少了拐角处约60%的毛刺。优化算法有效地修正了红外触摸屏原有的不足,能实现多触点时的轨迹关联和提高轨迹的平滑程度,能提高红外触摸屏绘图时的实际使用效果。     

基于Mean Shift迭代的新生未知多扩展目标跟踪

针对当前基于随机集的多扩展目标跟踪算法存在计算量大、量测划分不准确和跟踪误差大的问题,在PHD滤波基础上提出一种基于均值漂移(Mean Shift)迭代的新生未知多扩展目标跟踪算法.首先,对聚类后量测数据进行关联,得到新生目标状态,解决目标新生问题;然后,通过Mean Shift迭代获得目标量测集质心,将扩展目标的多量测问题转化为点量测处理;最后,给出其粒子实现方式.仿真实验表明,所提出的算法可以降低跟踪复杂度,提高跟踪效率,在交叉时刻具有稳定的跟踪性能。     

基于CKF的特征辅助数据关联多目标跟踪

针对密集杂波环境下的多目标近距跟踪问题,提出了一种基于容积卡尔曼滤波(CKF)和特征辅助数据关联的多目标跟踪算法(FADA-CKF).通过特征信息来对传统量测进行扩维,利用扩维后的量测对关联概率进行修正,将特征信息辅助技术融入到联合概率数据关联中,再利用容积卡尔曼滤波(CKF)处理非线性观测量,对目标状态进行估计.将FADA-CKF算法用于近距多目标跟踪场景中,仿真结果表明,改进算法在跟踪精度和误跟率方面要优于传统的JPDA跟踪算法.     

一种红外多目标跟踪的快速算法

含有新目标跟踪起始的数据关联问题是多目标跟踪算法研究中的一个难点,同时红外目标又有其自身的独特之处。首先描述了红外搜索跟踪(IRST)系统进行多目标跟踪中的跟踪起始和跟踪终结问题;然后提出了采用可变跟踪门的方法,并将IRST系统接收到的目标红外辐射作为一个参考量。给出了多目标跟踪起始和数据关联的快速算法。最后给出了数值仿真试验结果,仿真表明了这种算法的快速性和有效性。     

多传感器广义概率数据关联算法研究

随着跟踪环境、跟踪对象的不断变化与发展,目标与量测已经很难用一一对应的关系去描述;广义数据关联算法（GPDA）提出目标与量测多多对应的可行性规则,其性能在目标与量测无论是否在一一对应的情况下,均优于传统的JPDA算法,并且计算量、存储量均远小于JPDA;考虑到GPDA的上述优点,将GPDA算法拓展到多传感器数据关联,提出了多传感器广义概率数据关联算法（MSGPDA）来处理多传感器数据关联问题;仿真表明,MSGPDA算法由于利用多传感器信息,性能得到明显提高。     

基于联合概率数据关联算法的红外弱小目标跟踪

针对在云天背景下运动的红外弱小目标,从数据关联的角度,利用联合概率数据关联算法实现多个红外弱小目标的跟踪.实验结果表明,在杂波环境下,联合概率数据关联算法可以稳健地跟踪多个红外弱小目标的目标状态,跟踪效果好.     

雷达和红外融合对高机动目标跟踪仿真研究

研究准确跟踪敌机飞行目标,可进行有效打击.针对高机动目标与运动跟踪平台的相对运动比较复杂,由于雷达定位和红外跟踪算法的模型误差大与精度损失,如何提高跟踪系统的跟踪精度是一个很大的难题.为了提高高机动目标跟踪精确度,提出交互多模型-概率数据关联算法(IMM-PDA),进行雷达与红外并行信息融合与目标跟踪.运用Markov链切换原理与概率数据关联技术有效解决了目标变速机动及复杂杂波环境下的高精度跟踪问题.同时并行融合信息,克服序贯融合中两次使用扩展卡尔曼滤波线性化造成模型误差与精度损失的缺陷,提高系统的跟踪精度.仿真结果表明,基于雷达、红外并行融合的IMM - PDA算法在跟踪高机动目标时,精度更高,验证了算法的可行性与有效性.     

多传感器多目标联合概率数据关联研究

提出改进联合概率数据关联算法对多传感器、多目标量测进行同源划分及单一传感器测量数据转换,并采用联合概率数据关联算法求解空间目标轨迹交叉时的数据关联.仿真结果表明,改进联合概率数据关联算法提高了成功关联概率,降低了求解数据关联概率的难度,可以解决密集目标的正确跟踪问题.     

基于标签随机有限集的多量测多目标跟踪算法

针对一个扫描周期内单个目标可能产生多个量测的问题, 提出一种基于标签随机有限集的扩展算法. 结合脉冲扩展标签多伯努利(-GLMB) 滤波器和多量测模型, 推导出新的更新方程; 采用假设分解策略对关联过程进行降维, 避免了量测分组过程. 实验分析表明: 所提出算法能对目标数进行无偏估计, 在低探测概率条件下跟踪性能明显优于多量测概率假设密度(MD-PHD) 算法; 计算开销在量测较少时高于MD-PHD, 量测个数增加时增幅低于MD-PHD.

     

基于自适应变结构信息滤波的目标跟踪算法

目前目标跟踪算法采用的交互多模型,大多是通过固定模型之间的切换来完成目标跟踪,这容易出现模型集与目标真实运动不匹配问题,降低目标跟踪的精度。同时,现在大部分观测平台都能提供多传感器量测,这要求跟踪算法能对不同量测信息进行高效数据融合。针对上述问题,提出一种基于自适应变结构多模型和信息滤波的跟踪算法,它由少量模型构成模型集,通过在线更新模型集参数以自适应目标真实运动,采用无迹卡尔曼信息滤波融合多传感器量测信息,实现对目标的跟踪。仿真结果表明,该算法可以有效融合多传感器量测信息,自适应匹配目标真实运动,实现对目标稳定的高精度跟踪。     

基于雷达和红外传感器的目标检测方法

航迹关联是多传感器数据融合的一项重要内容,其目的在于评判不同的航迹是否来自同一个目标。取不同目标的位置和速度之间的距离作为模糊因子,提出了一种基于模糊推理的雷达和红外传感器航迹关联算法;该算法采用正态型隶属函数,在加权处理的基础上利用最大隶属函数原则进行航迹相似性评判。仿真结果表明了该算法的有效性。     

Variability of biome reflectance directional signatures as seen by POLDER

Reflectance measurements acquired with the spaceborne POLDER instrument are used to analyze the variability of land surface directional signatures as a function of vegetation cover type. The reflectance directional signatures are quantified by the three parameters of a modified version of the Ross-Li reflectance model. The variability of the estimated parameters with respect to the seven MODIS biome classes was found to be higher within the classes than between classes, with the exception of the desert targets that show more isotropic reflectances. A limited number of standard BRDFs (typically 5 in the red and near infrared) capture most of the variability of the directional reflectance measurements, supporting the idea that different land surfaces have similar directional signatures. Over vegetation targets, they are characterized by a strong increase toward backscattering and much smaller variations in forward directions. The results express the diversity in structural situations within a given biome class and indicate that, at the resolution of the POLDER sensor, i.e. a few kilometers, the BRDF contains little information on the dominant vegetation type. We show that standard directional signatures may be used to correct the reflectance measurements for directional effects with an RMS error on the order of 0.011 in the red and 0.015 in the near infrared.     

Sensor integration and data fusion

The problem of sensor integration and data fusion is addressed. We consider the problem of combining information from diversified sources in a coherent fashion. We assume that information from various sensors may be available in different forms at the fusion. For example, data from infrared (IR) sensors may be combined with range radar (RR) data and further combined with visual images. In each case, data and information from different sensors are presented in a different format which may not be directly compatible for all sensors. Part of the available information may be in the form of attributes and part in the form of dynamical measurements. A generalized evidence processing theory and an architecture for sensor integration and data fusion that accommodates diversified sources of information are presented. Data (or, more generically, information) fusion may take place at different levels, such as the level of dynamics, the level of attributes, and the level of evidence. The common and different aspects of fusion at the different levels are investigated and several practical examples of real world data fusion problems are discussed.     

Derivation of a shortwave infrared water stress index from MODIS near- and shortwave infrared data in a semiarid environment

Two different configurations of a shortwave infrared water stress index (SIWSI) are derived from the MODIS near- and shortwave infrared data. A large absorption by leaf water occurs in the shortwave infrared wavelengths (SWIR) and the reflectance from plants thereby is negatively related to leaf water content. Two configurations of a water stress index, SIWSI(6,2) and SIWSI(5,2) are derived on a daily basis from the MODIS satellite data using the information from the near infrared (NIR) channel 2 (841-876 nm) and the shortwave infrared channel 5 (1230-1250 nm) or 6 (1628-1652 nm), respectively, which are wavelength bands at which leaf water content influence the radiometric response. The indices are compared to in situ top layer soil moisture measurements from the semiarid Senegal 2001 and 2002, serving as an indicator of canopy water content. The year 2001 rainfall in the region was slightly below average and the results show a strong correlation between SIWSI and soil moisture. The SIWSI(6,2) performs slightly better than the SIWSI(5,2) (r²=0.87 and 0.79). The fieldwork in 2002 did not verify the results found in 2001. However, year 2002 was an extremely dry year and the vegetation cover apparently was too sparse to provide information on the canopy water content. To test the robustness of the SIWSI findings in 2001, soil moisture has been modelled from daily rainfall data at 10 sites in the central and northern part of Senegal. The correlations between SIWSI and simulated soil moisture are generally high with a median r²=0.72 for both configurations of the SIWSI. It is therefore suggested that the combined information from the MODIS near- and shortwave infrared wavelengths is useful as an indicator of canopy water stress in the semiarid Sahelian environment.     

Multi-sensor image super-resolution with fuzzy cluster by using multi-scale and multi-view sparse coding for infrared image

Super-resolution (SR) methods are effective for generating a high-resolution image from a single low-resolution image. However, four problems are observed in existing SR methods. (1) They cannot reconstruct many details from a low-resolution infrared image because infrared images always lack detailed information. (2) They cannot extract the desired information from images because they do not consider that images naturally come at different scales in many cases. (3) They fail to reveal different physical structures of low-resolution patch because they extract features from a single view. (4) They fail to extract all the different patterns because they use only one dictionary to represent all patterns. To overcome these problems, we propose a novel SR method for infrared images. First, we combine the information of high-resolution visible light images and low-resolution infrared images to improve the resolution of infrared images. Second, we use multiscale patches instead of fixed-size patches to represent infrared images more accurately. Third, we use different feature vectors rather than a single feature to represent infrared images. Finally, we divide training patches into several clusters, and multiple dictionaries are learned for each cluster to provide each patch with a more accurate dictionary. In the proposed method, clustering information for low-resolution patches is learnt by using fuzzy clustering theory. Experiments validate that the proposed method yields better results in terms of quantization and visual perception than the state-of-the-art algorithms.     

沉积岩石信息的热红外多光谱遥感探测及有效性评价

在对热红外多光谱（ＴＩＭＳ）遥感的探测机理及其图像专题信息提取方法的综合分析基础上，以不同沉积岩石的化学成分，矿物组成，结构，反射和发射光谱特征的实验室分析温度测试结果为依据，对不同沉积岩石的热红外多光谱遥感成像机制进行了探讨，并对比分析了不同沉积岩石ＴＩＭＳ图像和ＴＭ图像的识别效果。     

Spatial and temporal behaviour of a large agricultural area as observed from airborne C-band scatterometer and thermal infrared radiometer

Abstract

Natural thermal emission in the 8-12 μm band as well as emission or diffusion of microwaves at low frequencies are known to be strongly correlated with the soil surface water content. Theoretical studies of the interaction between the soil/plant system and the atmosphere have shown that such measurements may be used to monitor the soil water budget. Experimental results from an airborne campaign are presented here. The surface temperature and radar cross-section spatial properties and their interrelations are examined for very different situations (wet and dry, bare and vegetated). It is shown that, as far as large-scale applications are concerned, the field scale can be considered as homogeneous for both measurements and that it is possible to derive meaningful regional information from measurements at that scale. Therefore, at least for the test region, a high spatial resolution (of microwaves and thermal infrared measurements) is not required for the monitoring of surface soil moisture and thermal equilibrium for this 10-day period.     

基于信息扩散原理的关联规则自动评价方法研究 *

在关联规则的评价中 ,对评价参数的度量是参数在数据库上的平均分布 ,无法刻画参数在数据库的局部分布 ;另一方面 ,评价参数也是有限的。为解决上述问题 ,首先给出评价关联规则的五个参数及参数选取的原因分析 ;然后给出分库方案及其分库上的挖掘算法 ,通过分库挖掘算法获取各参数在分库上的参数值。对获取的参数值 ,利用扩散原理计算参数在数据库上的分布 ,通过分布实现对规则的评价和取舍 ,从而实现关联规则的自动评价和取舍 ,以及最大程度的自动知识获取 ,并具有实时查询特性。     

实现机器人随动的红外感知方法

结合机器人随动任务对人体运动定位的需求,提出一种分层递阶结构的红外运动感知模式,及基于热释电红外传感器的物理实现方法.在传感模型方面,采用两层递阶结构的感知模式:底层为由多个几何传感单元构成的几何传感层,在机器人视野内实现多视角运动目标的方位测量;上层为协作感知层,用来组织底层各几何传感单元的协作,实现运动目标的定位测量.在物理实现方面,底层组合热释电红外传感阵列与菲涅尔透镜组,构建运动目标方位角测量所需的几何传感单元;上层以最小二乘优化准则作为融合多视角方位测量信息的手段,获得运动目标的位置信息.移动机器人随动实验结果验证了所提出感知方法的有效性.较多视角光学视觉传感技术,该方法在传感效率、对光照和环境变化的稳健性、低成本和低功耗等方面具有不可替代的优势.     

基于语义分割的红外和可见光图像融合

红外图像即使在低光照条件下,也能根据热辐射的差异将目标与背景区分开来,而可见光图像具有高空间分辨率的纹理细节,此外,红外和可见光图像都含有相应的语义信息.因此,红外与可见光图像融合,需要既保留红外图像的辐射信息,也保留可见光图像的纹理细节,同时,也要反映出二者的语义信息.而语义分割可以将图像转换为带有语义的掩膜,提取源图像的语义信息.提出了一种基于语义分割的红外和可见光图像融合方法,能够克服现有融合方法不能针对性地提取不同区域特有信息的缺点.使用生成式对抗神经网络,并针对源图像的不同区域设计了2种不同的损失函数,以提高融合图像的质量.首先通过语义分割得到含有红外图像目标区域语义信息的掩模,并利用掩模将红外和可见光图像分割为红外图像目标区域、红外图像背景区域、可见光图像目标区域和可见光图像背景区域;然后对目标区域和背景区域分别采用不同的损失函数得到目标区域和背景区域的融合图像;最后将2幅融合图像结合起来得到最终融合图像.实验表明,融合结果目标区域对比度更高,背景区域纹理细节更丰富,提出的方法取得了较好的融合效果.