一种基于目标梯度特征和轨迹预测的复杂云背景目标检测算法

对于复杂云层背景下的红外弱小目标检测跟踪一直是图像处理的研究热点。受复杂云层和强光等因素的干扰,目标很容易淹没在背景中。针对传统方法无法兼顾较高的目标检测概率和较低的目标检测虚警率的问题,提出了一种目标检测跟踪方法。通过梯度预滤波的方法获取初始目标,二次比对的方法获取真实目标;通过目标航迹建立与删除的方法降低虚警率;通过抗云层遮挡卡尔曼滤波跟踪的方法提升目标检测能力。实验结果表明,该技术不仅很好地抑制了复杂云层背景和强光干扰背景,而且通过目标建航和抗遮挡跟踪等方法极大提升了目标检测概率,降低了虚警率。     

复杂红外地面环境下的稳定目标跟踪方法

针对红外目标相关滤波跟踪过程中由于背景杂波干扰、目标遮挡和目标形变等情况导致的鲁棒性差甚至跟踪目标丢失的问题,提出一种融合跟踪-学习-检测方法和相关滤波理论的红外目标跟踪算法.该算法在传统相关滤波框架基础上,融合目标的方向梯度直方图特征和亮度直方图特征,改善了目标轻微形变导致的模型漂移问题.针对背景杂波和遮挡导致的多峰值响应问题,对目标背景区域的相关响应进行惩罚,建立目标和背景响应的多模态检测机制,实现目标由粗到精的定位,并采用自适应的学习率优化跟踪模型的漂移问题;针对目标被严重遮挡或脱离视野的问题,通过全局目标再检测,实现目标的重捕.实验结果表明,在复杂红外地面环境下,该算法有效地解决了相似目标干扰和目标被严重遮挡导致的目标丢失问题.基于OTB-2015视频基准序列和红外视频序列测试,对比多个主流的相关滤波跟踪算法,该算法在跟踪精度和成功率方面较长时相关滤波跟踪算法分别提升了5.6%和4.1%;在目标遮挡指标测试中,该算法在跟踪精度和成功率方面相较长时相关滤波跟踪算法分别提升了4.6%和6.1%.     

基于加权mean-shift可见光/红外双通道目标跟踪

为提高目标亮度突变时的跟踪性能,在每一帧进行目标跟踪时,首先提取可见光图像的颜色特征,红外图像的垂直投影图像和水平投影图像特征,然后利用可见光/红外各自通道的Bhattacharyya系数计算该通道的权值,并对加权mean-shift双通道跟踪方法进行了推导,提出了基于加权mean-shift可见光/红外双通道目标跟踪算法.该方法使前后两帧目标相似度大的通道取大的权值,从而达到有效利用各通道有利信息、提高跟踪性能的目的.实验表明,用本文提出的算法进行可见光/红外双通道目标跟踪时,与基于mean-shift单通道(可见光或红外)目标跟踪算法相比,可提高目标跟踪的准确度,特别是当目标进入树荫区域,引起目标亮度发生显著变化时,跟踪性能基本不受影响.     

基于音视频信息融合的目标检测与跟踪算法

针对单一视觉跟踪算法易受遮挡影响的缺陷,提出一种基于音视频信息融合的目标检测与跟踪算法。整个算法框架包括视频检测与跟踪、声源定位、音视频信息融合跟踪3个模块。视频检测与跟踪模块采用YOLOv5m算法作为视觉检测的框架,使用无迹卡尔曼滤波和匈牙利算法实现多目标的跟踪与匹配;声源定位模块采用十字型麦克风阵列获取音频信息,结合各麦克风接收信号的时延计算声源方位;音视频信息融合跟踪模块构建音视频似然函数和音视频重要性采样函数,采用重要性粒子滤波作为音视频融合跟踪的算法,实现对目标的跟踪。在室内复杂环境下对算法性能进行测试,结果表明该算法跟踪准确率达到90.68%,相较于单一模态算法具有更好的性能。     

基于CamShift和Kalman组合的改进目标跟踪算法

针对应用CamShift算法进行目标跟踪过程中,当目标被严重遮挡、目标被与目标颜色相近的背景干扰时易丢失跟踪目标的问题,提出了一种基于CamShift和Kalman滤波组合的改进跟踪算法；为克服目标因严重遮挡而丢失的缺陷,利用自适应算法改进了传统的CamShift算法,扩大了搜索窗口,使运动目标位于搜索窗口内；为解决目标因颜色相近背景干扰而丢失的问题,改善跟踪准确率,利用卡尔曼滤波预测目标运动空间位置,作为下一帧搜索窗口的质心坐标；基于上述改进,利用C++语言,研发了改进的CamShift目标跟踪软件模块,给出了该模块的算法流程；实验结果表明,改进后的目标跟踪算法能有效地克服传统CamShift算法的缺陷,大大提高运动目标跟踪的准确性；所提的算法可以应用于运动小车跟踪,人脸识别等领域。     

基于稀疏编码直方图的稳健红外目标跟踪

充分利用红外图像信息建立有效的观测模型是实现稳健红外目标跟踪的基础。影响红外目标跟踪结果的因素除可见光目标跟踪也会面临的干扰因素之外,还有诸如边缘和纹理信息缺失、信噪比低和背景噪声影响等特有因素。提出基于稀疏编码直方图(HSC)特征和扰动感知模型(DAM)的红外目标跟踪方法,使用K-奇异值分解算法得到过完备字典,利用该字典计算得到每个像素点的稀疏编码,并组成HSC对目标进行表达,同时通过引入DAM增强算法抗背景干扰能力。该方法充分利用了红外图像中运动目标的结构特性,能够有效去除背景干扰。与其他跟踪器相比,在VOT-TIR2015数据集上,该方法的精确度和成功率指标分别获得3.8%和4.4%的提升,具有较高的研究价值和实用价值。     

基于多级跟踪队列的运动目标跟踪遮挡处理

针对多目标跟踪过程中存在的遮挡问题,提出了一种固定摄像机场景下的多目标实时跟踪算法.提出基于鬼影判别与背景模型选择更新的背景差法检测运动目标,建立一种融合色度与边缘特征的目标模型.通过定义稳定跟踪队列、临时跟踪队列、跟踪丢失队列以及候选跟踪队列等跟踪器队列,提出基于多级关联匹配的策略实现多目标跟踪遮挡处理,针对新目标、...     

基于图像融合的动态轮廓线跟踪新方法

红外与可见光传感器是目标跟踪识别系统中常用的两种传感器，对这两种传感器图像进行融合能有效提高系统跟踪检测的准确性。将动态轮廓线模型与图像融合结合，在特征搜索过程中利用特征点准确地完成了图像配准，同时使用了一种新的特征级融合方法，将两种图像中目标轮廓的B样条曲线控制点进行实时微分耦合。这种耦合将Curwen提出的微分耦合机制作了改进，利用图像配准把刚性硬模板改变为实时的变换模板并推导了融合后动态轮廓线的新的动力学方程。这种融合利用了红外图像目标轮廓信息约束可见光图像中动态轮廓线的收敛形状，有效地提高了可见光图像目标跟踪的准确性。对运动人手序列图像的对比跟踪实验表明，这种融合使得可见光图像中动态轮廓线平均跟踪误差减小了60．25％。     

基于形变多样相似性的空间正则化相关滤波跟踪

提出了一个基于形变多样相似性的空间正则化相关滤波跟踪算法。在核相关滤波(KCF)跟踪算法基础上引入了空间正则化权重和子网格检测方法,利用形变多样相似性匹配算法构建了目标重检测模块,利用主成分分析(PCA)算法和k维树一致近似最近邻(TreeCANN)算法解决了匹配算法中的最近邻搜索问题;通过自适应模板更新策略,解决了遮挡情况下模板误更新问题。实验结果表明,所提算法的精确度得分为0.825,成功率得分为0.625,相比KCF算法分别提升了18.5%和31.0%。所提算法能较好地解决目标尺度变化、遮挡、快速运动、旋转和背景杂乱情况下的跟踪问题,具有广泛的应用前景。     

基于全局感知孪生网络的红外目标跟踪

目前大多数热红外(TIR)目标跟踪算法都是基于相关滤波或者使用彩色跟踪器的模型进行特征提取。然而,两者都存在适用于彩色目标跟踪却对红外目标特征不敏感的缺陷,导致无法良好地应用到红外目标跟踪。为此,提出一种基于全局感知的孪生神经网络的红外目标跟踪器。将孪生神经网络的后三层特征进行融合优化,得到新的特征,同时加入了由空间转换网络和通道注意力组成的空间感知模块,以得到全局范围内的有效信息,通过引入自注意力机制,使算法更加专注于提取目标的判别信息,最后对结果进行响应融合得到最终的响应图。在PTB-TIR红外目标跟踪评估基准上的实验结果表明,本文算法能够适应多样的红外环境,同时能够保持良好的跟踪速度(20.2 frame/s),实现对红外目标有效且稳定的实时跟踪。     

相位一致性图像及其在目标跟踪中的应用(英文)

针对传统实时相关跟踪方法对照度变化敏感的问题,提出了一种基于相位一致性图像的相关跟踪方法.利用相位一致性函数值在[0,1]区间内且无量纲、对图像的亮度和对比度具有不变性等特点,首先对原始图像进行相位一致性检测,得到相位一致性图像,再利用MAD(Minimum Absolute Difference)等相关跟踪算法在相位一致性图像中对目标进行跟踪运算.对可见光和红外图像的实验表明,在图像的亮度和对比度发生剧烈变化的情况下,算法仍能保持对目标的稳定跟踪.该方法可用于解决传统实时相关跟踪方法普遍存在的因照度变化导致跟踪点漂移甚至跟踪失败的问题.     

Object tracking in a stereo and infrared vision system

In this paper, we deal with the problem of real-time detection, recognition and tracking of moving objects in open and unknown environments using an infrared (IR) and visible vision system. A thermo-camera and two stereo visible-cameras synchronized are used to acquire multi-source information: three-dimensional data about target geometry and its thermal information are combined to improve the robustness of the tracking procedure. Firstly, target detection is performed by extracting its characteristic features from the images and then by storing the computed parameters on a specific database; secondly, the tracking task is carried on using two different computational approaches. A Hierarchical Artificial Neural Network (HANN) is used during active tracking for the recognition of the actual target, while, when partial occlusions or masking occur, a database retrieval method is used to support the search of the correct target followed. A prototype has been tested on case studies regarding the identification and tracking of animals moving at night in an open environment, and the surveillance of known scenes for unauthorized access control.     

基于高光谱成像技术的微小摄像头检测技术

针对高光谱成像特点,提出了一种基于三维特征检测微小摄像头的方案。在空间维利用猫眼效应筛选疑似目标,在光谱维对结果进行精准判定。依据摄像头结构,分析了可见光摄像头的反射光谱特征。基于几何光学和辐射度学,计算和仿真了系统的探测距离。结果表明,正常工作时,光功率影响最小探测距离,目标尺寸影响最大探测距离。搭建了微小摄像头光谱特征验证系统。结果表明,采用吸收型红外截止滤光片的目标的非反射光占比曲线变化平缓且数值高,采用反射型红外截止滤光片的目标的非反射光占比曲线可见光部分数值高,红外部分数值低,从700 nm附近开始下降,甚至发生突变,实验数据显示,突变位置的斜率绝对值是红外波段斜率绝对值的10倍以上。实验结果与预期分析的结果一致,验证了高光谱成像技术检测微小摄像头的可行性。     

Visible and infrared tracking based on multi-view multi-kernel fusion model

In the visual tracking problem, fusion of visible and infrared sensors provides complementarily useful features and can consistently help distinguish the target from the background efficiently. Recently, multi-view learning has received growing attention due to its enormous potential in combining diverse view features containing consistent and complementary characteristics. Therefore, in this paper, a visible and infrared fusion tracking algorithm based on multi-view multi-kernel fusion (MVMKF) model is presented. The proposed MVMKF model considers the diversities of visible and infrared views and embeds complementary information from them. Furthermore, the multi-kernel framework is used to learn the importance of view features so that an integrated appearance representation is made with regard to the respective performance. Besides, the tracking task is completed with naive Bayes classifier in sophisticated compressive feature domain, considering the high performances of classifier-level and sophisticated feature-level learning for multiple views. The experimental results demonstrate that the MVMKF tracking algorithm performs well in terms of accuracy and robustness.     

一种基于算法融合的红外目标跟踪方法

视频目标跟踪的难点在于快速、准确地在帧与帧之间匹配目标.由于红外图像目标与背景的反差低,图像的边缘模糊并且灰度级动态范围小,使红外目标跟踪难度比可见光更大.本文提出一种针对红外日标跟踪的融合算法,该方法融合直方图和不变矩的特点.首先利用目标的直方图计算简单快速的特点,由均值平移算法快速找到局部最优解,但由于该局部最优解仪为直方图匹配的最优解,缺少目标形状特征,与实际目标位置存在一定的偏差;其次,利用边缘小变矩作为修正特征修正误差,避免跟踪误差逐渐累计而最终导致跟踪失败,以提高跟踪的稳定性和精度.实验结果表明,该算法能够消除跟踪过程中的漂移现象,提高跟踪精度.     

Joint detection and tracking of size-varying infrared targets based on block-wise sparse decomposition

The high variability of target size makes small target detection in Infrared Search and Track (IRST) a challenging task. A joint detection and tracking method based on block-wise sparse decomposition is proposed to address this problem. For detection, the infrared image is divided into overlapped blocks, and each block is weighted on the local image complexity and target existence probabilities. Target-background decomposition is solved by block-wise inexact augmented Lagrange multipliers. For tracking, label multi-Bernoulli (LMB) tracker tracks multiple targets taking the result of single-frame detection as input, and provides corresponding target existence probabilities for detection. Unlike fixed-size methods, the proposed method can accommodate size-varying targets, due to no special assumption for the size and shape of small targets. Because of exact decomposition, classical target measurements are extended and additional direction information is provided to improve tracking performance. The experimental results show that the proposed method can effectively suppress background clutters, detect and track size-varying targets in infrared images.     

自动对靶喷雾靶标红外探测研究

农林作物一般非连续种植,喷在农林作物靶标之间的药液造成巨大浪费和严重环境污染,因此探讨利用红外线探测技术实现靶标的自动探测,对靶精准喷施意义重大.红外线可有效防止可见光的干扰,响应速度快,可实现无接触探测,由此建立的红外光电探测系统可靠性好、成本低.红外发光管发射出红外光照射到被探测物体上,反射的红外线被光电探测器接收,触发控制信号,实现自动对靶施药.系统中不同探测器组采用不同编码的调制脉冲红外信号,可消除组间光路干扰及其他光信号的干扰.实验结果表明所建立的红外靶标自动探测装置可对农林作物靶标进行自动探测,光波波长为850 nm,探测距离为0.1～6 m可调,靶标最小识别间距小于0.3 m.     

A new infrared small and dim target detection algorithm based on multi-directional composite window

This work presents a new method based on gray characteristic analysis for infrared dim small target detection under complex backgrounds. Firstly, an improved detection window with eight directions and three layers is introduced to investigate the gray distribution characteristic of different structure in an infrared image. Secondly, we adopt a pretreatment process based on morphology filter and mean filter to reduce the running time and propose a detection rule on characteristic analysis for infrared targets. Meanwhile a new parameter optimization algorithm based on fuzzy control theory is employed so that the detection rule could be independent of the initial parameters. Finally, experimental results indicate that the proposed method can effectively detect the dim small targets and has better tracking performance.     

Detection and tracking of targets in infrared images using Bayesian techniques

Simple yet robust techniques for detecting targets in infrared (IR) images are an important component of automatic target recognition (ATR) systems. In our previous works, we have developed IR target detection and tracking algorithms based on image correlation and intensity. In this paper, we discuss these algorithms, their performances and problems associated with them and then propose novel algorithms to alleviate these problems. Our proposed target detection and tracking algorithms are based on frequency domain correlation and Bayesian probabilistic techniques, respectively. The proposed algorithms are found to be suitable for real-time detection and tracking of static or moving targets, while accommodating for detrimental affects posed by the clutter and background noise. Finally, limitations of all these algorithms are discussed.