几种自动目标跟踪算法的比较研究

复杂背景目标跟踪是近年来自动目标识别(ATR)领域的一个研究热点,在军事、医疗、安全等多个领域具有广泛的应用前景。ATR的研究内容主要包括目标的检测分类、特征提取和目标定位识别等。本文对当前流行的目标跟踪算法进行了全面的分析比较,最后对目标跟踪算法的进一步研究方向进行了深入的探讨。     

基于多核DSP的红外目标识别算法

近年来红外传感器在军事、安防领域得到广泛应用,基于嵌入式平台的红外目标识别算法是该领域的重点研究方向。该文提出了一种基于梯度方向特征的景象匹配算法,其可利用可见光模板对红外图像进行异源图像间的目标识别。该文阐述了在算法移植过程中,如何基于TI公司的TMS320C6678 DSP芯片底层硬件、指令级流水线设计,对算法及算法框架进行优化设计和编程,该方法可广泛适用于其他图像处理算法在DSP上的移植优化。     

基于分形特征和导引滤波的可见光与红外图像融合算法

为充分利用红外图像的目标指示信息,本文提出了一种基于分形特征和导引滤波的可见光与红外图像融合算法。该算法首先采用分形特征对红外图像中的人造目标进行增强,通过阈值分割得到目标分布图。待融合图像经过一层分解得到近似图像和细节图像,基于目标分布图利用导引滤波分别得到可见光与红外近似图像与细节图像的融合系数。实验结果表明,融合后图像充分结合了可见光图像的背景信息与红外图像中的目标信息,有利于后续的目标识别任务。     

改进最大类别方差法的电力设备视觉识别技术研究

针对变电站电力设备的视觉识别红外图像普遍存在信噪比和对比度相对较低的问题,笔者在多种红外图像分割方法的基础上,提出了一种改进的最大类间方差法,通过实验证明该方法在变电站应用的普适性,并将改进的分割方法与模式匹配识别算法相结合,应用到电力设备红外图像目标识别方向,提高了电力设备红外图像识别技术的识别精度和决策性能。     

利用模糊综合评判技术提取红外图像海天线

海天线检测是红外图像自动目标识别技术的一项重要内容.本文深入分析了海天背景红外图像的特点,提出了一种基于四项特征评分因子的海天线模糊识别算法.首先对梯度图进行Radon变换以提取候选海天线,然后联合候选海天线四种特征信息进行模糊综合评判,最后根据评判值大小给出海天线识别结果.算法对比和实测数据结果表明,该方法适用范围广,抗噪声和抗干扰能力强,识别率高,可以有效地检测出各种复杂海天背景下的海天线,为进一步应用(如目标识别)奠定了良好的基础.     

核最小二乘算法检测红外点目标(英文)

对于背景呈非线性变化的复杂图像,用背景预测的方法对红外点目标进行检测时,传统的线性最小二乘法(Least Squares,LS)的效果比较差.文章使用核方法(Kernel Methods,KMs)推导了最小二乘法的非线性版本:核最小二乘算法(Kernel Least Squares,KLS);进一步推导出了更适合动态系统时序预测的指数加权形式的核最小二乘算法(Kemel Exponential wleighted Least Squares,KEWLS).提出了一种基于核方法的红外点目标检测算法,先用KEWLS非线性回归算法预测红外图像背景,再通过自适应门限检测残差图像中的目标,非线性函数回归和红外序列图像检测实验表明核方法较大地改进了算法的非线性函数估计与红外背景预测能力.     

自动目标识别(综述)

本文综述了解决自动目标识别(ATR)问题的各种技术,包括目标检测、分割、特征计算、分类、跟踪等。重点放在目前主要采用的模式识别(PR)方法,同时,对新的方法,例如以知识为基础(K—B)方法,和多传感器ATR问题,也给予评述。     

红外小目标的证据理论识别方法

提出了一种红外序列图像中小目标的Dempster-Shafer(DS)证据理论识别方法。DS证据理论在图像融合中存在两个实际问题:一是经典的DS证据理论组合算子在处理高冲突信息时会产生违反常理的结果,二是如何将图像特征转化为证据理论中各个命题的基本概率指派。针对这两个问题,引入证据距离,改进证据组合公式,解决了高冲突证据下的信息融合问题,保证了融合算法的可靠性;运用模糊逻辑,求出了图像中各像素隶属于目标的基本概率指派。对实际的红外序列图像运用改进的公式进行像素级融合,并根据最大概率准则确定出图像中的小目标。实验结果表明,新方法更为有效、可靠,识别错误率可降低到6.14%。     

海空背景下电影经纬仪中目标红外图像特性分析

针对海空复杂背景下红外小目标识别的需要,对海空背景及其弱小目标的红外图像特性进行了统计分析。根据图像特性的差异,可以判断图像中是否存在目标,从而为目标识别提供了关于目标识别和背景的先验知识,研究结果可用于海空背景下小目标的检测。     

红外弱小目标检测背景抑制算法研究

复杂背景下红外小目标识别一直是红外图像处理的关键技术之一,针对复杂云背景下红外弱小目标的时域和空域特征,考虑到易于硬件实现和实时性要求,提出基于快速统计排序滤波和Robinson Guard滤波并行快速处理算法,对复杂背景进行高信噪比抑制。实验证明,该方法能够有效地提高红外弱小目标图像信噪比和复杂背景下的小目标的检测概率。     

Automatic target recognition employing signal compression

Quadratic correlation filters (QCFs) have been used successfully to detect and recognize targets embedded in background clutter. Recently, a QCF called the Rayleigh quotient quadratic correlation filter (RQQCF) was formulated for automatic target recognition (ATR) in IR imagery. Using training images from target and clutter classes, the RQQCF explicitly maximized a class separation metric. What we believe to be a novel approach is presented for ATR that synthesizes the RQQCF using compressed images. The proposed approach considerably reduces the computational complexity and storage requirements while retaining the high recognition accuracy of the original RQQCF technique. The advantages of the proposed scheme are illustrated using sample results obtained from experiments on IR imagery.     

用数学形态学方法进行红外图像分层解理

将形态学算子作用于高噪声,低质量的红外图像,对之进行分层解理还原,即将原本由各种信号叠加而成的红外图像重新分解成器件噪声,工作噪声和目标,背景等不同层次的信号。这种图像解理方法,以几何论和集合运算为基础,因而原理直观,计算简单,能为红外目标识别与跟踪提供有益的预处理。     

一种基于压缩视频的红外图像预处理算法

文章讨论了在DCT域中进行红外视频图像预处理的可能性,并提出了一种基于DCT域的压缩视频预处理算法.该算法通过DCT域低通滤波、系数变换和对比度拉伸实现了红外图像的增强处理.仿真结果表明,提出的算法克服了对视频压缩、解压再处理的弊端,提高了系统的整体性能和响应速度,改善了视频图像的视觉效果.     

Methods to detect objects in photon-limited images

We investigate the problem of detecting and localizing a known signal in a photon-limited image, where Poisson noise is the dominant source of image degradation. For this purpose we developed and evaluated three new algorithms. The first two are based on the impulse restoration (IR) principle and the third is based on the generalized likelihood ratio test (GLRT). In the IR approach, the problem is formulated as one of restoring a delta function at the location of the desired object. In the GLRT approach, which is a well-known variation on the optimal likelihood ratio test, the problem is formulated as a hypothesis testing problem, in which the unknown background intensity of the image and the intensity scale of the object are obtained by maximum-likelihood estimation. We used Monte Carlo simulations and localization receiver operating characteristic (LROC) curves to evaluate the proposed algorithms quantitatively. LROC curves demonstrate the ability of an algorithm to detect and locate objects in a scene correctly. Our simulations demonstrate that the GLRT approach is superior to all other tested algorithms.     

高分辨率遥感图像飞机目标定位新算法

为了提高自动目标识别系统的准确性,本文提出了一种高分辨率遥感图像的飞机目标定位的新方法,即在图像分割后再进行一次"去除伪目标"的目标识别系统新策略.首先,利用先验知识采用重叠的面积均分法对图像分块,然后基于多次阈值自动分割方法对各块图像进行分割.针对大图像目标识别的分块策略和飞机目标分割不完全的实际情况,提出一种基于辅助直线的Radon变换飞机轴线检测方法,用于飞机目标的定位.最后,通过实验证明,当飞机目标分割不完全时,本文方法能有效的检测飞机目标.     

Attenuated total reflection-Fourier transform infrared imaging of large areas using inverted prism crystals and combining imaging and mapping

Attenuated total reflection-Fourier transform infrared (ATR-FT-IR) imaging is a very useful tool for capturing chemical images of various materials due to the simple sample preparation and the ability to measure wet samples or samples in an aqueous environment. However, the size of the array detector used for image acquisition is often limited and there is usually a trade off between spatial resolution and the field of view (FOV). The combination of mapping and imaging can be used to acquire images with a larger FOV without sacrificing spatial resolution. Previous attempts have demonstrated this using an infrared microscope and a Germanium hemispherical ATR crystal to achieve images of up to 2.5 mm x 2.5 mm but with varying spatial resolution and depth of penetration across the imaged area. In this paper, we demonstrate a combination of mapping and imaging with a different approach using an external optics housing for large ATR accessories and inverted ATR prisms to achieve ATR-FT-IR images with a large FOV and reasonable spatial resolution. The results have shown that a FOV of 10 mm x 14 mm can be obtained with a spatial resolution of approximately 40-60 microm when using an accessory that gives no magnification. A FOV of 1.3 mm x 1.3 mm can be obtained with spatial resolution of approximately 15-20 microm when using a diamond ATR imaging accessory with 4x magnification. No significant change in image quality such as spatial resolution or depth of penetration has been observed across the whole FOV with this method and the measurement time was approximately 15 minutes for an image consisting of 16 image tiles.     

Integrated approach for automatic target recognition using a network of collaborative sensors

We introduce what is believed to be a novel concept by which several sensors with automatic target recognition (ATR) capability collaborate to recognize objects. Such an approach would be suitable for netted systems in which the sensors and platforms can coordinate to optimize end-to-end performance. We use correlation filtering techniques to facilitate the development of the concept, although other ATR algorithms may be easily substituted. Essentially, a self-configuring geometry of netted platforms is proposed that positions the sensors optimally with respect to each other, and takes into account the interactions among the sensor, the recognition algorithms, and the classes of the objects to be recognized. We show how such a paradigm optimizes overall performance, and illustrate the collaborative ATR scheme for recognizing targets in synthetic aperture radar imagery by using viewing position as a sensor parameter.     

红外图像的动态阈值分割

在深入探讨目前图像分割算法的基础上，针对实际红外图像的分割提出了一种新的方法－－动态阈值分割算法。它利用局部空间阈值适度调整全局统计最优阈值，因而其分割阈值是动态的。实验表明该算法能够突出图像中感兴趣的细节，单动态阈值的分割效果相当于常规固定双域值的分割。     

基于互信息和混合算法的印刷图像配准

目的为了提高印刷图像配准的精度,提出一种基于混合搜索算法的图像配准方法。方法首先求取图像的归一化互信息,然后利用GA算法(遗传算法)进行全局搜索,得出粗配准参数;最后,利用Powell算法进行局部寻优,得出精配准参数。结果混合算法的配准结果与只用单一Powell搜索算法或只用单一GA搜索算法相比,在各个几何变换方向上得到了更小的配准误差。结论与GA算法和Powell算法相比,文中建议的混合算法配准精确度更高、速度更快。