Background subtraction based on time-series clustering and statistical modeling

This paper proposes a robust method to detect and extract silhouettes of foreground objects from a video sequence of a static camera based on the improved background subtraction technique. The proposed method analyses statistically the pixel history as time series observations. The proposed method presents a robust technique to detect motions based on kernel density estimation. Two consecutive stages of the k-means clustering algorithm are utilized to identify the most reliable background regions and decrease the detection of false positives. Pixel and object based updating mechanism for the background model is presented to cope with challenges like gradual and sudden illumination changes, ghost appearance, non-stationary background objects, and moving objects that remain stable for more than the half of the training period. Experimental results show the efficiency and the robustness of the proposed method to detect and extract the silhouettes of moving objects in outdoor and indoor environments compared with conventional methods.     

Moving object detection based on improved VIBE and graph cut optimization

Extracting foreground moving objects from video sequences is an important task and also a hot topic in computer vision and image processing. Segmentation results can be used in many object-based video applications such as object-based video coding, content-based video retrieval, intelligent video surveillance and video-based human–computer interaction. In this paper, we present a novel moving object detection method based on improved VIBE and graph cut method from monocular video sequences. Firstly, perform moving object detection for the current frame based on improved VIBE method to extract the background and foreground information; then obtain the clusters of foreground and background respectively using mean shift clustering on the background and foreground information; Third, initialize the S/T Network with corresponding image pixels as nodes (except S/T node); calculate the data and smoothness term of graph; finally, use max flow/minimum cut to segmentation S/T network to extract the motion objects. Experimental results on indoor and outdoor videos demonstrate the efficiency of our proposed method.     

多目标跟踪中目标数量准确识别算法

由于场景中存在目标遮挡与重叠的影响,以及检测时产生的分割错误,同一个目标往往被识别为若干运动区域,或者多个目标被识别为同一个运动区域,从而导致目标数量识别错误。为了解决这一问题,提出了一种结合融合与分离操作的理论来分析视频的方法,将多目标跟踪的问题转化成为一个找寻后验极值的问题。采用图论的方法表示观测结果,利用图像序列中目标运动轨迹和外观相似度的信息,将多目标跟踪的问题归结为找寻图中多个最优路径的问题。该方法采用了滑动窗口框架以便统计固定数目帧的信息。实验结果表明该方法能够应对实际中发生的上述现象,达到了准确识别多目标数量的目的。     

New color correction method for multi-view video using disparity vector information

Color inconsistency is an urgent problem to be solved in free viewpoint television. In this letter, a new color correction method is proposed by using disparity vector information. At first, we separate foreground and background from the scene with a method of mean-removed disparity estimation. Then the correction parameters are estimated by adopting linear fitting for foreground and background regions, respectively. Next, with expectation-maximization algorithm, we integrate correction parameters of foreground and background to get the final corrected image. Finally, video tracking technique is performed to correct multi-view video. Experimental results show that the proposed method is quite effective.     

基于自适应阈值的齿轮干涉图像前景区域提取方法

齿面物体像灰度法是激光移相干涉测量中提取齿轮干涉图像前景区域的重要方法之一,针对该方法因人工设定阈值且忽略不同图像边缘特征从而导致的测量效率及精度受限问题,提出了一种基于自适应阈值的齿轮干涉图像前景区域提取方法。首先分析齿轮齿面形貌特征与各边缘顶点差异,对图像进行区域划分;然后根据边缘灰度变化规律通过邻域窗口筛选合格像素点并获取掩模结果,实现前景区域提取;最后根据5类图像评价指标分别对4组算法分割结果与传统方法分割结果进行数据对比。结果表明：算法在实现图像自动处理的基础上与参考结果匹配精确度提升约3.5%～4.5%,PRI(probabilistic rand index)提升约3%～4%,VOI(variation of information)提高约15%～25%,GCE(global consistency error)降低约2.5%～3.5%,最终相位信息准确度提升9μm～15μm。结果符合精度要求,该方法可广泛应用于齿轮干涉图像前景提取中。     

A self-adaptive optical flow method for the moving object detection in the video sequences

This paper proposes a self-adaptive optical flow method to detect moving objects in the video sequences. The method first estimates the original optical flow field with the optical flow algorithm, and then enhances the objects by a local mean algorithm, and finally filters out the noise with a self-adaptive threshold algorithm. The proposed method has a wide adaptivity to the size and the number of objects, and it also can effectively process the scenarios of complex background and that of the slight occlusion. Furthermore, it avoids the complicated and time-consuming preprocessing procedure. The results of the present method show that the moving objects can be detected effectively.     

动态背景下基于光流场分析的运动目标检测算法

针对现有动态背景下运动目标检测算法的不足,提出一种基于光流场分析的运动目标检测算法.首先根据前背景在光流梯度幅值和光流矢量方向上的差异确定目标的大致边界,然后通过点在多边形内部原理获得边界内部的稀疏像素点,最后以超像素为节点,利用混合高斯模型拟合的表观信息和超像素的时空邻域关系构建马尔可夫随机场模型的能量函数,并通过使目标函数能量最小化得到最终的运动目标检测结果.该算法不需要任何先验假设,能够同时处理动态背景和静态背景两种情况.多组实验结果表明,本文算法在检测的准确性和处理速度上均优于现有算法.     

基于阈值判断的CamShift目标跟踪算法

针对CamShift算法只利用目标的颜色信息,在跟踪过程中,易受目标相似物、遮挡以及光照等复杂背景影响导致目标搜索窗口发散,跟踪稳定性能降低,提出了一种基于阈值判断的目标跟踪方法。该方法将OTSU法和Snake模型结合,利用OTSU法以最佳阈值对图像进行分割,分离前景区域和背景区域,初步提取目标轮廓作为Snake模型的初始轮廓,经收敛得到目标的精准轮廓,利用轮廓外接最小矩形框内的像素计算目标质心,判断与CamShift算法中目标搜索窗口质心之间的欧式距离,如果未超出阈值,则直接使用CamShift算法跟踪目标,反之,则将计算出的目标质心作为CamShift算法中当前帧目标搜索窗口的质心跟踪目标。实验结果表明,该算法跟踪目标具有较好的实时性,跟踪性能稳定、可靠。     

基于特征点光流聚类的复杂背景中运动车辆检测

为了准确、快速的在动态场景中对运动车辆进行检测,提出一种基于特征点光流聚类的车辆检测方法。该方法取Harris角点为特征量,通过对特征点做光流提取来剔除一些没有运动的干扰角点,然后再通过模糊U邻域(FUNN)聚类算法剔除噪音、孤立点和不感兴趣样本并实现前景和背景的分离,最后通过设定阈值判断前景目标是否是车辆。实验结果证明在复杂的动态场景中该算法具有更高的车辆识别率。     

基于随机聚类的复杂背景建模与前景检测算法

为了构建鲁棒的背景模型和提高前景目标检测的准确性, 综合考虑同一位置的像素点在时间上的关联性和与其相邻像素的空间关联性, 基于经典的ViBe算法中的随机聚类思想提出了一种复杂背景建模和前景检测方法. 利用样本一致性原理, 采用前n帧序列图像得到初始化背景, 避免了Ghost现象的发生; 根据实际复杂背景的动态反馈获取自适应聚类阈值和自适应更新阈值进行随机聚类, 从而实现了对动态背景的适应性; 通过全局扰动阈值和局部像素级判断阈值的结合, 实现了对光照缓慢变化、快速变化以及突然变化的免疫性, 准确地分割前景目标. 对多组数据集的测试结果表明, 本文算法较大地提高了背景模型对动态背景、光照变化及相机抖动的复杂背景的适应性和鲁棒性. 算法还能很好地适用于红外图像检测运动目标的场合, 扩展了本算法的应用范围. 在没有进行任何图像预处理和形态学后处理情况下, 得到的原始前景检测精度优于其他对比算法.     

图-谱结合的压缩感知高光谱视频图像复原

针对液晶可调滤波片高光谱成像系统记录动态场景的成像特点,提出一种图-谱结合的压缩感知高光谱视频图像复原方法。首先,通过前景目标检测获得运动前景目标的高光谱图像,实现运动前景目标与背景区域分离,并根据前景目标检测结果将背景区域划分为运动区域(被前景目标遮挡区域)与静止区域(未被前景目标遮挡区域)。然后,基于高光谱图像空间维、光谱维相关性,对静止区域进行字典学习获得稀疏先验信息,结合压缩感知理论用于运动区域恢复,得到完整的背景区域高光谱图像。最后,将运动前景目标高光谱图像与背景区域高光谱图像相结合,得到高光谱视频图像。实验结果表明:本文提出的高光谱视频图像复原方法在峰值信噪比和视觉效果上都要优于现有算法,峰值信噪比平均提高5 d B以上。     

Singular Spectrum Analysis for Background Initialization with Spatio-Temporal RGB Color Channel Data

In video processing, background initialization aims to obtain a scene without foreground objects. Recently, the background initialization problem has attracted the attention of researchers because of its real-world applications, such as video segmentation, computational photography, video surveillance, etc. However, the background initialization problem is still challenging because of the complex variations in illumination, intermittent motion, camera jitter, shadow, etc. This paper proposes a novel and effective background initialization method using singular spectrum analysis. Firstly, we extract the video’s color frames and split them into RGB color channels. Next, RGB color channels of the video are saved as color channel spatio-temporal data. After decomposing the color channel spatio-temporal data by singular spectrum analysis, we obtain the stable and dynamic components using different eigentriple groups. Our study indicates that the stable component contains a background image and the dynamic component includes the foreground image. Finally, the color background image is reconstructed by merging RGB color channel images obtained by reshaping the stable component data. Experimental results on the public scene background initialization databases show that our proposed method achieves a good color background image compared with state-of-the-art methods.     

Fisher’s linear discriminant ratio based threshold for moving human detection in thermal video

In video surveillance, the moving human detection in thermal video is a critical phase that filters out redundant information to extract relevant information. The moving object detection is applied on thermal video because it penetrate challenging problems such as dynamic issues of background and illumination variation. In this work, we have proposed a new background subtraction method using Fisher’s linear discriminant ratio based threshold. This threshold is investigated automatically during run-time for each pixel of every sequential frame. Automatically means to avoid the involvement of external source such as programmer or user for threshold selection. This threshold provides better pixel classification at run-time. This method handles problems generated due to multiple behavior of background more accurately using Fisher’s ratio. It maximizes the separation between object pixel and the background pixel. To check the efficacy, the performance of this work is observed in terms of various parameters depicted in analysis. The experimental results and their analysis demonstrated better performance of proposed method against considered peer methods.     

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.     

Adaptive threshold selection for background removal in fringe projection profilometry

In fringe projection profilometry, background and shadow are inevitable in the image of an object, and must be identified and removed. In existing methods, it is nontrivial to determine a proper threshold to segment the background and shadow regions, especially when the gray-level histogram of the image is close to unimodal, and an improper threshold generally results in misclassification of the object and the background/shadow. In this paper, an adaptive threshold method is proposed to tackle the problem. Different from the existing automatic methods, the modulation-level histogram, instead of the gray-level histogram, of the image is employed to determine the threshold. Furthermore, a new weighting factor is proposed to improve Otsu's method to segment the image with a histogram close to unimodal, and the modulation difference of the object pixels and the background/shadow pixels is intensified significantly by the weighting factor. Moreover, the weighting factor is adaptive to the image. The proposed method outperforms existing methods either in accuracy, efficiency or automation. Experimental results are given to demonstrate the feasibility and effectiveness of the proposed method.     

面向对象分析的非结构化背景目标高光谱探测方法研究

针对非结构化背景探测器中背景协方差矩阵估计的局限性,提出了一种基于面向对象分析的高光谱小目标探测算法。首先对图像进行自适应迭代分割处理,将其划分为许多均质对象;然后进行正态最优分布选取,利用多元正态无偏检验选取最佳对象集;最后将此数据集合作为局部背景并结合GLR基准算法进行目标探测。该算法可以使局部背景最大化的服从正态分布,有效地将背景光谱信息和目标光谱信息分离开来,同时通过最优选取过程克服了目标信息“污染”问题。为了验证算法的有效性,利用真实的OMIS数据进行仿真实验,并与非结构化背景探测器GLR和基于K-Means聚类的改进GLR算法的检测结果比较,结果表明提出的算法具有良好的探测性能和较低的虚警概率。     

LTP与光照突变补偿结合的运动目标检测算法

针对现有运动目标检测算法在光照突变条件下鲁棒性不强、易发生误检等问题,提出了将LTP算子与光照突变补偿模型结合的运动目标检测算法。首先利用LTP算子获取当前帧与其背景图像的纹理特征图,然后计算当前帧图像像素点作为前景点的概率并依据概率自适应地更新背景,再判断图像是否发生光照突变补偿并采用线性模型补偿突变图像序列,最后利用背景减除法获取运动目标。实验结果表明本文算法在光照突变条件下对运动目标检测效果理想,检测精度指标PR高于同类其他算法,且具有良好的鲁棒性。     

基于感兴趣区域的幅型比变换新方法

基于人类视觉系统对运动目标比较敏感的特性,提出了一种新的幅型比变换方法.综合时空信息提取视频序列中的运动目标,利用图像修复技术对被提取运动目标后的视频图像进行背景修补.采用插值算法以不同的方式分别缩放运动目标和修复后的背景,并将变换后的运动目标与背景相叠加合成相应幅型比的视频图像.实验结果表明：此方法避免了已有幅型比变换方法的缺陷,保证了变换后运动目标的无失真性,与人类的视觉感知有较好的一致性.     

Application of piecewise linear regression twice to inspect the defective regions on thin film transistor liquid crystal display panels

Detecting defects in thin film transistor liquid crystal display (TFT-LCD) panels can be difficult due to non-uniform background brightness and slightly different brightness levels between the defective regions and the background. One well-known method is to inspect the defects of TFT-LCD panels using the polynomial approximation method. However, there are some problems such as determination of the polynomial degree and ghost objects. In this paper, to overcome these problems, we propose a new algorithm using the piecewise linear regression twice and Niblack’s method which do not require to determine the polynomial degree for the approximation. Our algorithm effectively removed ghost objects. The experimental results show that our proposed method performed well and was good enough to substitute the polynomial approximation method.