Moving object detection method based on complementary multi resolution background models

A novel moving object detection method was proposed in order to adapt the difficulties caused by intermittent object motion,thermal and dynamic background sequences.Two groups of complementary Gaussian mixture models were used.The ghost and real static object could be classified by comparing the similarity of the edge images further.In each group,the multi resolution Gaussian mixture models were used and dual thresholds were applied in every resolution in order to get a complete object mask without much noise.The computational color model was also used to depress illustration variations and light shadows.The proposed method was verified by the public test sequences provided by the IEEE Change Detection Workshop and compared with three state-of-the-art methods.Experimental results demonstrate that the proposed method is better than others for all of the evaluation parameters in intermittent object motion sequences.Four and two in the seven evaluation parameters are better than the others in thermal and dynamic background sequences,respectively.The proposed method shows a relatively good performance,especially for the intermittent object motion sequences.

An Improved Moving Object Detection Algorithm Based on Gaussian Mixture Models

Aiming at the problems that the classical Gaussian mixture model is unable to detect the complete moving object, and is sensitive to the light mutation scenes and so on, an improved algorithm is proposed for moving object detection based on Gaussian mixture model and three-frame difference method. In the process of extracting the moving region, the improved three-frame difference method uses the dynamic segmentation threshold and edge detection technology, and it is first used to solve the problems such as the illumination mutation and the discontinuity of the target edge. Then, a new adaptive selection strategy of the number of Gaussian distributions is introduced to reduce the processing time and improve accuracy of detection. Finally, HSV color space is used to remove shadow regions, and the whole moving object is detected. Experimental results show that the proposed algorithm can detect moving objects in various situations effectively.

The Discrete Representation of Continuously Moving Indeterminate Objects

To incorporate indeterminacy in spatio-temporal database systems, grey modeling method is used for the calculations of the discrete models of indeterminate two dimension continuously moving objects. The Grey Model GM（ 1,1 ） model generated from the snapshot sequence reduces the randomness of discrete snapshot and generates the holistic measure of object＇s movements. Comparisons to traditional linear models show that when information is limited this model can be used in the interpolation and near future prediction of uncertain continuously moving spatio-temporal objects.

Grain Yield Prediction of Henan Province Based on Spatio-temporal Regression Model

By using correlation analysis method,regression analysis method and time sequence method,we combine time and space,to establish grain yield spatio-temporal regression prediction model of Henan Province and all prefecture-level cities.At first,we use the grain yield in prefecture-level cities of Henan in the year 2000 and 2005,to establish regression model,and then taking the grain yield in one year as independent variable,we predict the grain yield in the fifth year afterwards.Taking the dependent variable value as independent variable again,we predict the grain yield at an interval of the same years,and based on this,predict year by year forward until the year we need.The research shows that the grain yield of Henan Province in the year 2015 and 2020 is 59.849 6 and 67.929 3 million t respectively,consistent with the research results of other scholars to some extent.

Real-time moving object detection for video monitoring systems

Moving object detection is one of the challenging problems in video monitoring systems, especially when the illumination changes and shadow exists. Amethod for real-time moving object detection is described. Anew background model is proposed to handle the illumination varition problem. With optical flow technology and background subtraction, a moving object is extracted quickly and accurately. An effective shadow elimination algorithm based on color features is used to refine the moving obj ects. Experimental results demonstrate that the proposed method can update the background exactly and quickly along with the varition of illumination, and the shadow can be eliminated effectively. The proposed algorithm is a real-time one which the foundation for further object recognition and understanding of video mum＇toting systems.

GSM-MRF based classification approach for real-time moving object detection

Statistical and contextual information are typically used to detect moving regions in image sequences for a fixed camera.In this paper,we propose a fast and stable linear discriminant approach based on Gaussian Single Model(GSM)and Markov Random Field(MRF).The performance of GSM is analyzed first,and then two main improvements corresponding to the drawbacks of GSM are proposed:the latest filtered data based update scheme of the background model and the linear classification judgment rule based on spatial-temporal feature specified by MRF.Experimental results show that the proposed method runs more rapidly and accurately when compared with other methods.

LayeredModel:一个面向室内空间的移动对象数据模型

基于室内空间的移动对象管理或称室内移动对象管理,是一个崭新而又富有挑战的研究领域.如何建立语义完备并且支持多种应用的室内移动对象模型是一个重要的基础性问题.提出了一个分层的室内移动对象数据模型:LayeredModel.该模型提出了室内空间距离的概念,并使用不同的层次关系表达室内元素、传感器以及移动对象之间的关系.从而能支持多种基于室内空间的应用,例如移动对象的跟踪、监控、导航以及室内最近邻查询等.LayeredModel模型也为后续的室内移动对象的索引、查询的研究奠定了基础.

基于改进ViBe的自适应运动目标检测算法

针对ViBe算法无法去除动态背景,易出现鬼影及不能自适应光照变化的问题,提出一种复杂环境自适应的ViBe改进算法。通过计算区域的复杂度、闪烁波动度,对分类半径R和更新率T进行动态调整,对样本点进行有效性权重的计算,更高效地过滤背景噪声和适应光照渐变;在检测物体状态变化时,动态调整R和T,通过融合前景点计数和帧差法优化鬼影消除;通过识别最小外接矩阵区域差异加快去除鬼影;利用帧差法实时检测光照突变,及时进行重新初始化,避免大量误检。实验结果表明,改进ViBe算法在适应动态背景、光照变化及抑制鬼影等方面比原算法均有更好检测效果,检测精度平均提升了40.7%。

基于超宽带技术的运动目标跟踪高精度定位系统设计

为克服运动目标不断变化导致跟踪定位精度较低的问题,设计基于超宽带技术的运动目标跟踪高精度定位系统;图像采集模块由FPGA单元、VGA显示单元、帧缓存单元以及图像采集单元构成,以此实现运动目标跟踪与定位中的图像采集,在超宽带技术模块中,设计超宽带运动目标定位所需的天线、定位基站、移动节点,完成超宽带动态组网,实现硬件系统的设计;基于硬件系统采集到的图像,实施图像灰度化处理、形态学滤波处理,以增强图像中有用的信息,设计TLD运动目标跟踪算法,随着运动目标开始运动,TLD模型会不断学习跟踪的运动目标,获取目标在距离、景深、角度等层面的改变,并不断学习、识别,达到良好的跟踪效果,基于超宽带技术设计运动目标动态定位算法,依据跟踪结果实现运动目标的高精度定位,完成软件系统的设计;实验测试结果表明,该系统在中、远距离目标跟踪与定位实验中跟踪错误率低于0.60%、2.4%,沿着S型运动时,路线弯折处的定位误差较低,与实验运动目标的飞行路线相贴合,具有良好的定位能力。

基于卡尔曼(Kalman)滤波算法的体育训练视频中的运动目标检测

体育训练视频中的运动目标检测往往存在目标快速移动、目标遮挡和场景变化等问题,导致运动目标检测的难度增大。为了解决这些问题,提出了基于卡尔曼(Kalman)滤波算法的体育训练视频中运动目标检测的方法。为了提升体育训练视频图像的质量,首先对体育训练视频的图像进行一系列预处理(包括灰度变换、轮廓对比增强和噪声抑制等);然后借助混合高斯模型(Gaussian mixture module,GMM)有效提取体育训练视频的前景信息;为了精准捕捉体育训练视频中的运动目标,运用三帧差分法设定Kalman滤波器的初始状态,利用高效检测算法准确获取每一帧图像中运动目标的观测位置;随后将初始状态和观测位置的数据输入Kalman滤波器;最后在Kalman滤波器中,结合上一帧的预估值和当前帧的监测值,对体育训练视频中运动目标的当前状态进行精确估算和优化,并对下一帧的状态进行预测,从而实现了对运动目标的持续跟踪与精准预测。实验结果表明,与基于特征融合的全卷积孪生网络(siamese full convolution,Siamfc)目标追踪算法和基于连续自适应均值漂移(continuously adapting mean shift,Camshift)的均值漂移(mean shift,Meanshift)改进算法相比,采用基于Kalman滤波算法的运动目标检测方法对运动目标的重叠精度(overlap precision,OP)和中心位置误差(center location error,CLE)进行检测,不仅能够有效检测到体育训练视频中的运动目标,还可表现出较高的准确性和实时性。

Video Frame’s Background Modeling: Reviewing the Techniques

Background modeling is a technique for extracting moving objects in video frames. This technique can be used in ma-chine vision applications, such as video frame compression and monitoring. To model the background in video frames, initially, a model of scene background is constructed, then the current frame is subtracted from the background. Even-tually, the difference determines the moving objects. This paper evaluates a number of existing background modeling techniques in term of accuracy, speed and memory requirement.

Neural network based method for background modeling and detecting moving objects

This paper proposes a novel method, primarily based on the fuzzy adaptive resonance theory （ART） neural network with forgetting procedure, for moving object detection and background modeling in natural scenes. With the ability, inheriting from the ART neural network, of extracting patterns from arbitrary sequences, the background model based on the proposed method can learn new scenes quickly and accurately. To guarantee that a long-life model can derived from the proposed mothed, a forgetting procedure is employed to find the neuron that needs to be discarded and reconstructed, and the finding procedure is based on a neural network which can find the extreme value quickly. The results of a suite of quantitative and qualitative experiments conducted verify that for processes of modeling background and detecting moving objects our method is more effective than five other proven methods with which it is compared.

基于改进ViBe算法的运动目标检测

ViBe算法是一种基于静态背景下的运动目标检测算法,针对其“鬼影”问题和运动目标静止时会被更新为背景的问题提出了改进ViBe算法,即对原ViBe算法的背景模型初始化、动态阈值、前景分割和背景模型更新等4个部分进行了改进。采用均值法获取的背景图像初始化背景模型,可消除“鬼影”;利用计数法控制前景分割动态阈值,使前景图像更加准确;使用帧差法思想改进前景分割,使前景图像更加完整;通过引入阈值保证背景模型更新的稳定性。根据试验结果可知,改进ViBe算法对正常移动车辆、较小运动目标和存在静止情况的运动目标都有较好的检测能力,解决了“鬼影”问题和运动目标静止时会被更新为背景的问题,同时相较于原ViBe算法和其他常用运动目标检测算法,改进ViBe算法在保证准确性的基础上提高了检测的完整性。

基于可靠性低秩因子分解和泛化差异性差分的运动目标检测

运动目标检测旨在分离视频的背景与前景,然而常用的低秩因子分解法往往难以综合地处理动态背景和间歇性运动的问题。考虑到背景减除后的偏态噪声分布具有潜在的背景修正作用,提出一种基于可靠性低秩因子分解和泛化差异性差分的运动目标检测模型。首先,利用时间维度像素分布的峰值位置以及偏态分布性质选取一个不含离群像素的子序列,并计算该子序列的中值以形成静态背景;其次,利用非对称拉普拉斯分布对静态背景减除后的噪声建模,并把基于空间平滑的建模结果作为可靠性权重参与到低秩因子分解中,以此建模综合背景(含有动态背景);最后,依次利用时间和空间连续约束提取前景。其中,针对时间连续性,提出了泛化差异性差分约束,从而通过相邻视频帧的差异信息抑制前景边缘的扩增。实验结果表明,与PCP、DECOLOR、LSD、TVRPCA、E-LSD、GSTO六种模型相比,所提模型的F-measure值最高。由此可知,所提模型在动态背景、间歇性运动等复杂场景中能有效提高前景的检测精度。

Adaptive Indexing of Moving Objects with Highly Variable Update Frequencies

In recent years, management of moving objects has emerged as an active topic of spatial access methods. Various data structures （indexes） have been proposed to handle queries of moving points, for example, the well-known B^x-tree uses a novel mapping mechanism to reduce the index update costs. However, almost all the existing indexes for predictive queries are not applicable in certain circumstances when the update frequencies of moving objects become highly variable and when the system needs to balance the performance of updates and queries. In this paper, we introduce two kinds of novel indexes, named B^y-tree and αB^y-tree. By associating a prediction life period with every moving object, the proposed indexes are applicable in the environments with highly variable update frequencies. In addition, the αB^y-tree can balance the performance of updates and queries depending on a balance parameter. Experimental results show that the B^y-tree and αB^y-tree outperform the B^x-tree in various conditions.

AKAZE与高斯混合模型相融合的运动目标检测算法

传统高斯混合模型在抖动干扰下无法取得理想的运动目标检测效果,但实际的各种复杂情况下摄像系统均会存在抖动。针对这一问题,文中提出一种抖动干扰下运动目标检测算法:首先利用基于AKAZE特征点匹配算法,进行动态图像精确配准,消除图像抖动,做出运动补偿;其次利用高斯混合模型对配准后的视频序列进行运动目标检测,然后利用三帧差分技术对检测出的运动目标轮廓进一步补偿;最后使用连通域分析补全运动目标。实验表明,本文提出的方法能够满足检测要求,可以在抖动干扰下有效检测出运动目标。

基于快速背景建模的垃圾车识别系统

为保障城市道路交通安全、实现绿色环保出行,城市道路智能监控愈发重要,重点监控车辆如城市垃圾车的智能识别也成为其中关键一环。为提高城市道路智能监控系统分析能力,设计了基于快速背景建模的垃圾车车型识别系统。首先,设计了垃圾车识别系统的网络拓扑结构,并构建了车辆视频侧视采集系统以保留丰富的车辆特征。其次,为降低复杂背景的影响,提出背景更新控制策略,实现了快速可靠的背景建模;为实现侧视车辆图像的有效分割,提出了基于动态位置检测的关键帧筛选方法和基于斜线投影的目标分割方法。最后,设计了基于SVM(Support Vector Machine)的多级分类器实现了垃圾车识别。试验结果表明,文中方法垃圾车识别的平均准确率为95%以上、平均召回率为90%以上。

Moving Objects with Transportation Modes:A Survey

In this article,we survey the main achievements of moving objects with transportation modes that span the past decade.As an important kind of human behavior,transportation modes reflect characteristic movement features and enrich the mobility with informative knowledge.We make explicit comparisons with closely related work that investigates moving objects by incorporating into location-dependent semantics and descriptive attributes.An exhaustive survey is offered by considering the following aspects:1)modeling and representing mobility data with motion modes;2)answering spatio-temporal queries with transportation modes;3)query optimization techniques;4)predicting transportation modes from sensor data,e.g.,GPS-enabled devices.Several new and emergent issues concerning transportation modes are proposed for future research.

Bayesian moving object detection in dynamic scenes using an adaptive foreground model

Accurate detection of moving objects is an important step in stable tracking or recognition. By using a nonparametric density estimation method over a joint domain-range representation of image pixels, the correlation between neighboring pixels can be used to achieve high levels of detection accuracy in the presence of dynamic background. However, color similarity between foreground and background will cause many foreground pixels to be misclassified. In this paper, an adaptive foreground model is exploited to detect moving objects in dynamic scenes. The foreground model provides an effective description of foreground by adaptively combining the temporal persistence and spatial coherence of moving objects. Building on the advantages of MAP-MRF (the maximum a posteriori in the Markov random field) decision framework, the proposed method performs well in addressing the challenging problem of missed detection caused by similarity in color between foreground and background pixels. Experimental results on real dynamic scenes show that the proposed method is robust and efficient.