基于改进巴氏指标和模型更新的视觉跟踪算法

传统的Mean Shift算法采用巴氏系数度量模型与候选模型之间的统计特征相似性,但是由于背景特征的影响,有时应用巴氏指标进行匹配得到最优解的位置并不一定是目标的实际位置,在跟踪过程中可能导致目标定位出现偏差。该文提出一种改进的巴氏系数相似度指标,指标由于引入了前景/背景置信值,能够有效抑制待匹配区域中背景特征的影响,突出目标特征的权重,与原始的巴氏指标相比,明显提高了目标匹配的准确性。基于改进的巴氏指标,对目标与背景区域双模型相似度系数进行综合分析,合理地判断干扰目标匹配的原因,从而采取相应的模型更新策略。采用4段具有挑战性的视频序列对5种跟踪算法进行了测试,通过定量实验分析可知,文中算法处理1帧视频所需的平均时间为75.76 ms,实时性仅次于原始的Mean Shift跟踪算法,同时跟踪误差在5种跟踪算法中取得了最优结果。实验结果表明,该算法能够有效抑制背景干扰和避免模型漂移,在不同的复杂场景下都具有一定的鲁棒性。     

An occlusion-adaptive tracker based on sparse representation using alternating direction method of multipliers

Recently, sparse representation has been applied to visual tracking with satisfactory performance. However, partial occlusion and computational complexity are two main obstructions in developing sparse-based tracking. In this paper, a simple yet robust tracker based on patch-based sparse representation is proposed. An adaptive motion model, including adaptive sampling regions and adaptive particle numbers, is proposed to improve the sampling efficiency. A self-adjustable segmentation approach is proposed to segment the target into local patches. A patch-based observation model, which is occlusion-adaptive, is constructed by solving a set of L1-regularized least squares problems. The L1-regularized least squares problem is solved using the alternating direction method of multipliers (ADMM). Both quantitative and qualitative experiments are conducted on several challenging image sequences and the comparisons with several state-of-the-art trackers demonstrate the effectiveness and efficiency of our tracker.     

Indirect target detection method in FLIR image sequences

Due to the complexity of the scene, target detection in forward-looking infrared (FLIR) imagery is a challenging problem, especially for occluded target. The main contribution of this paper is to propose an indirect detection method for improving the recognition probability and effectiveness of target detection method in FLIR image sequences under complex conditions. The proposed method mainly includes four steps: preparation of forward-looking reference image of landmark, extraction of the real-time scene image, template matching and target location, in which some key technologies are proposed, such as perspective transformation used to solve projective problems, position prediction for improving real-time performance, and target location used for identifying the target’s position. Experimental results are shown to demonstrate the robustness and efficiency of proposed method in FLIR image sequences.     

复杂背景下红外图像弱小目标检测

红外弱小目标检测是安防监控、侦察探测、精确制导等领域的关键技术。为了提高复杂背景条件下红外弱小目标检测的准确性和实时性,提出了一种基于深度学习的红外弱小目标检测算法YOLO-FCSP。根据红外图像中弱小目标的特点,在YOLO检测框架的基础上,通过减少下采样次数,结合跨阶段局部模块、Focus结构和空间金字塔池化结构设计了特征提取网络。借鉴多路径聚合的思路优化特征融合网络,同时调整检测输出层数量,通过信息复用提高特征利用效率。实验结果表明,本文提出的算法在检测红外弱小目标时具有较高的准确率和检测速度,精度和召回率分别为91.9%和94.6%,平均准确率(AP)值达到92.6%,检测速度达到170 f/s,满足实际应用中实时检测的需求。     

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.     

Infrared moving object detection based on local saliency and sparse representation

The key issue of infrared object detection is to locate moving object in image sequence. In order to improve detection precision, an infrared object detection method based on local saliency and sparse representation is proposed in this paper. Motion information, such as velocity, acceleration components are added into the eigenvectors to build local saliency model. And the approximate position of the infrared target is located based on the local saliency. To accurately extract the infrared object, sparse representation is used to capture complete edge of the object. Experiments show that the proposed method can accurately detect infrared moving objects, and has good robustness to external disturbances and dynamic background.     

Optical superimposition of infrared thermography through video projection

This paper presents a novel infrared thermography visualization technique where a sequence of captured thermal images is optically and simultaneously superimposed onto the target object via video projection in real time. In conventional thermography visualization, observers have to frequently move their eyes from the object to a 2D screen where a thermal image is displayed. In contrast, the heat distribution of the object’s surface emerges directly onto its physical surface in the proposed method. As a result, the observer can intuitively understand the object’s heat information just by looking at it in the real space. This paper explains the methods of geometric registration and radiometric compensation of the captured thermal image, which are required before video projection. Furthermore, several projection results are shown to validate the intuitiveness and usefulness of the proposed visualization method.     

复杂红外背景中运动小目标快速跟踪技术

在地面、海面、天空复杂背景下对红外小目标稳定跟踪是亟需解决的难题.为兼顾鲁棒性和实时性,以判别尺度空间跟踪算法为基础,应用可有效表征目标区域灰度突变特性和目标形状信息的广义结构张量算法作为特征提取方法.改进后的算法更适用于红外图像快速处理,能提高跟踪鲁棒性,且计算量小、效率高,目标特征维度低.为提高跟踪稳定性,依据置信...     

强激光与红外光学系统光轴平行性检测方法的探讨

介绍3种可测量强激光发射光轴与红外光轴平行性的方法及每种方法的优缺点，重点研究了利用热靶进行波段转换来测量激光与红外光学系统光轴平行性的方法。该方法是将激光聚焦在热靶上，使热靶产生热量并发出红外光，红外光再经过准直进入被测设备的红外光学系统，从而测量出激光光轴与红外光轴的平行性。热靶材料的选择与激光透过率的确定尤为重要，该方法中选用酚醛树脂作为热靶材料，激光的透过率仅为0.5%。通过与远距离目标靶测量法进行比对实验，发现2种方法得到的测量结果一致，从而验证了这种方法的可行性与正确性。     

基于热红外图像处理技术的农作物冠层识别方法研究

为解决农作物冠层热红外图像边缘灰度级分布不均且噪声较大,而传统图像分割方法难以实现其目标区域有效识别的难题,以苗期红小豆冠层热红外图像为研究对象,将模糊神经网络和仿射变换有机结合,提出了基于热红外图像处理技术的农作物冠层识别模型。首先利用五层线性归一化模糊神经网络的自适应特性,选取高斯隶属度函数,自动计算冠层可见光图像识别的推理规则,有效地分割了可见光图像中的冠层区域。通过分析3种分割指标和熵,定量评价可见光图像冠层分割质量。网络迭代38次时,误差精度为0.000 952,该算法平均有效识别率为96.13%,获取可见光冠层图像的像元信息熵值范围为2.454 4~5.198 7,与标准算法所得冠层图像的像元信息熵仅相差0.245 9。然后以取得可见光图像的冠层有效区域为参考图像,采用仿射变换算法,调整优选平移、旋转、缩放等图像变换因子,配准原始热红外图像,提出了基于仿射变换的冠层热红外图像识别方法。对于初始温度范围值在16.35~19.92 ℃的农作物热红外图像,计算选取旋转幅度为1.0和缩放因子为0.9时,作为异源图像的最优配准参数,获取目标图像的最大温差为3.17 ℃,相对于原图像的平均温度值由18.711 ℃下降至17.790 ℃,进而实现了基于热红外图像处理技术的农作物冠层识别。最后以熵的互信息作为监督指标,对农作物冠层热红外图像识别方法进行评价。提出的冠层热红外图像识别方法,所获取的目标图像与初始热红外图像的平均互信息为4.368 7,标准目标图像和初始热红外图像的平均互信息为3.981 8,二者仅相差0.486 9。同时,两种冠层热红外图像的平均温度差值为0.25 ℃,高效消除了原始热红外图像的背景噪声。结果表明本研究方法的有效性和实用性,能够为应用热红外图像反映农作物生理生态信息特征指标参数提供技术借鉴。     

基于特征自适应选择的金字塔均值漂移跟踪方法

针对均值漂移跟踪算法框架不足以对目标帧间运动过大及快速尺度变化进行有效地处理,且单个图像特征对环境适应性较差.提出了一种特征自适应选择方法,通过分析目标与背景的特征区分度来选择出最有效的特征.将金字塔自适应分解和均值漂移跟踪结合,提出了金字塔均值漂移跟踪方法.采用背景加权直方图描述目标模板模型,核函数加权直方图描述候选...     

Fusion of infrared polarization and intensity images using support value transform and fuzzy combination rules

Infrared polarization and intensity imagery provide complementary and discriminative information in image understanding and interpretation. In this paper, a novel fusion method is proposed by effectively merging the information with various combination rules. It makes use of both low-frequency and high-frequency images components from support value transform (SVT), and applies fuzzy logic in the combination process. Images (both infrared polarization and intensity images) to be fused are firstly decomposed into low-frequency component images and support value image sequences by the SVT. Then the low-frequency component images are combined using a fuzzy combination rule blending three sub-combination methods of (1) region feature maximum, (2) region feature weighting average, and (3) pixel value maximum; and the support value image sequences are merged using a fuzzy combination rule fusing two sub-combination methods of (1) pixel energy maximum and (2) region feature weighting. With the variables of two newly defined features, i.e. the low-frequency difference feature for low-frequency component images and the support-value difference feature for support value image sequences, trapezoidal membership functions are proposed and developed in tuning the fuzzy fusion process. Finally the fused image is obtained by inverse SVT operations. Experimental results of visual inspection and quantitative evaluation both indicate the superiority of the proposed method to its counterparts in image fusion of infrared polarization and intensity images.     

Hybrid multi-resolution detection of moving targets in infrared imagery

A hybrid moving target detection approach in multi-resolution framework for thermal infrared imagery is presented. Background subtraction and optical flow methods are widely used to detect moving targets. However, each method has some pros and cons which limits the performance. Conventional background subtraction is affected by dynamic noise and partial extraction of targets. Fast independent component analysis based background subtraction is efficient for target detection in infrared image sequences; however the noise increases for small targets. Well known motion detection method is optical flow. Still the method produces partial detection for low textured images and also computationally expensive due to gradient calculation for each pixel location. The synergistic approach of conventional background subtraction, fast independent component analysis and optical flow methods at different resolutions provide promising detection of targets with reduced time complexity. The dynamic background noise is compensated by the background update. The methodology is validated with benchmark infrared image datasets as well as experimentally generated infrared image sequences of moving targets in the field under various conditions of varying illumination, ambience temperature and the distance of the target from the sensor location. The significant value of F-measure validates the efficiency of the proposed methodology with high confidence of detection and low false alarms.     

Frequency modulated infrared imaging for thermal characterization of nanomaterials

The paper presents frequency modulated thermal wave imaging (FMTWI) as a fast and efficient non-contact technique for in-plane thermal characterization of thin plate nanomaterials. A novel excitation signal in the form of an up-chirp is applied and the thermal response is monitored using an infrared (IR) thermography based temperature sensing system. The in-plane thermal diffusivity of any sample can be measured using the multiple phase information extracted from a single run of the experiment. This feature provides a time efficient approach for thermal measurements using infrared thermography techniques. The theoretical background and experimental details of the technique are discussed, with practical measurement of thermal diffusivity of an empty anodic alumina (AAO) template in direction perpendicular to the nanochannel axis, in support.     

Adaptive contour-based statistical background subtraction method for moving target detection in infrared video sequences

A robust contour-based statistical background subtraction method for detection of non-uniform thermal targets in infrared imagery is presented. The foremost step of the method comprises of generation of background frame using statistical information of an initial set of frames not containing any targets. The generated background frame is made adaptive by continuously updating the background using the motion information of the scene. The background subtraction method followed by a clutter rejection stage ensure the detection of foreground objects. The next step comprises of detection of contours and distinguishing the target boundaries from the noisy background. This is achieved by using the Canny edge detector that extracts the contours followed by a k-means clustering approach to differentiate the object contour from the background contours. The post processing step comprises of morphological edge linking approach to close any broken contours and finally flood fill is performed to generate the silhouettes of moving targets. This method is validated on infrared video data consisting of a variety of moving targets. Experimental results demonstrate a high detection rate with minimal false alarms establishing the robustness of the proposed method.     

A novel compression algorithm for infrared thermal image sequence based on K-means method

High resolution in space and time is becoming the new trend of thermographic inspection of equipments, therefore, the development of a fast and precise processing and data store technique of high resolution thermal image should be well studied. This article will propose a novel global compression algorithm, which will provide an effective way to improve the precision and processing speed of thermal image data. This new algorithm is based on the decay of the temperature of thermograph and the feature of thermal image morphology. Firstly, it will sort the data in space according to K-means method. Then it will employ classic fitting calculation to fit all the typical temperature decay curves. At last, it will use the fitting parameters of the curves as the parameters for compression and reconstruction of thermal image sequence to achieve the method for which the thermal image sequence can be compressed in space and time simultaneously. To validate the proposed new algorithm, the authors used two embedded defective specimens made of different materials to do the experiment. The results show that the proposed infrared thermal image sequence compression processing algorithm is an effective solution with high speed and high precision. Compared to the conventional method, the global compression algorithm is not only noise resistant but also can improve the computing speed in hundreds of times.     

Pattern recognition in hyperspectral imagery using one-dimensional shifted phase-encoded joint transform correlation

Pattern recognition in hyperspectral imagery is a challenging issue because of the high false alarm rate and computation complexity. In this paper, a one-dimensional shifted phase-encoded fringe-adjusted joint transform correlation (SPFJTC) technique is developed for hyperspectral image processing system. The proposed technique processes the reference spectral signature using a random phase mask and correlates it with the spectral signature corresponding to each pixel of the unknown input hyperspectral image cube using a simple architecture. This technique generates very high discrimination between the object of interest and background clutter. Computer simulation results using real life hyperspectral imagery show that the proposed SPFJTC technique can effectively recognize the objects of interest while alleviating the effects of false alarms and other artifacts.     

基于颜色传递的双波段目标跟踪算法研究

针对复杂背景下Camshift算法容易丢失目标的情况,提出一种基于YCBCR空间将红外与可见光融合图像与彩色参考图像进行颜色传递后,采用Camshift进行目标跟踪的算法。该算法在颜色传递时充分利用双波段图像信息,得到的目标对比度高并且颜色空间较其周围背景突出,增强了目标的颜色概率图,提高了Camshift算法效率。实验表明,通过对可见光图像、经颜色传递后的红外图像以及传统颜色传递方法得到的图像采用相同跟踪算法进行定性分析,在该算法得到的图像中,跟踪窗口中心相对于目标质心仅有3个像素的误差,跟踪精度远远优于对比实验图像的跟踪结果,并且算法的跟踪时间为每帧20.6 ms,达到了实时性的要求。     

成像目标跟踪目标建模技术综述

由目标跟踪的数学模型得出,影响目标跟踪性能的三个主要因素为目标状态转移模型、滤波算法和目标建模技术.对目标建模技术进行了综述和分析,分别从特征选择、特征的统计建模和相似性度量三个方面进行了阐述.以畸变不变性、目标/背景分辨能力作为性能评价手段定性地比较了国内外文献中提出的多种目标表征模型.指出了目标跟踪中目标表征模型自...