空间约束混合高斯运动目标检测

目的 针对传统混合高斯模型前景检测运算量过大问题,提出一种基于空间约束的混合高斯前景检测算法。方法 通过快速初始化缩短模型的初始建立过程;采用双重背景模型机制,以自适应背景减法的前景检测结果作为混合高斯前景检测的空间约束条件,降低模型在背景区域的冗余运算;运用多策略自适应模型更新,提高前景检测的准确性。结果 在各种测试场景下,与传统混合高斯法、CodeBook、GMG、偏差均值混合高斯模型(MODGMM)等算法相比,该算法具有更好的准确率以及4倍以上的处理速度。结论 在固定相机场景下的运动目标检测中,算法能有效提高传统混合高斯法的准确性且具有极高的实时性。     

一种基于自适应混合模型的背景减除法

针对局部二进制相似度(LBSP)背景建模方法易受外界环境变化如动态背景、光照改变、相机抖动等干扰的问题,在融合像素纹理与亮度信息的基础上,建立一种自适应混合背景模型进行运动目标检测。首先,利用每个像素的多通道自适应局部二进制相似度(LBSP)信息和亮度信息建立混合背景模型。然后,根据当前像素与混合背景模型的比较结果对其进行分类,并采用随机更新机制更新背景模型。实验结果表明,本方法不仅在正常外界环境下取得了较好的检测结果,而且还可以有效地减少动态背景、光照变化等复杂外界环境条件造成的干扰,提高检测结果的准确性。     

Exploring and modeling unimanual object manipulation on multi-touch displays

Touch-sensitive devices are becoming increasingly wide-spread, and consequently gestural interfaces have become familiar to the public. Despite the fact that many gestures require frequently dragging, pinching, spreading, and rotating the finger-tips, there currently does not exist a human performance model describing this interaction. In this paper, a novel user performance model is derived for virtual object manipulation on touch-sensitive displays, which involves simultaneous translation, rotation, and scaling of the object. Two controlled experiments with dual-finger unimanual manipulations were conducted to validate the new model. The results indicate that the model fits the experimental data well (with R² and R values above 0.9), and performs the best among several alternative models. Moreover, based on the analysis of the empirical data, the simultaneity nature of manipulation in the task is explored and several design implications are provided.     

双馈风力发电机控制系统半实物仿真平台设计

针对目前风力发电机仿真仅限于某个功能或单元的建模及分析,难以进行整机控制系统仿真的问题,以双馈发电机为对象,设计与实现了风力发电机半实物仿真实验平台.该平台包括控制系统、风力机虚拟对象、通讯接口以及监控界面组成,软件之间通过自动化设备规范(ADS)协议进行数据交互融合,为整机及控制系统的性能在线评估提供支持.最后在所开发实验平台上进行了低风速仿真实验,验证了所设计平台的有效性和合理性,为风电场后期的参数评估及性能优化提供一个完善可行的实用化平台.     

动态背景下的运动目标检测

针对动态背景下运动目标检测的问题,最大限度地降低背景对运动目标检测的影响,提出了一种基于相位相关法和傅里叶梅林变换的动态背景下运动目标检测算法.动态背景下运动目标检测的主要部分是背景运动补偿,首先利用相位相关法和傅里叶梅林变换估计全局运动参量,然后根据全局运动参量利用双线性内插法进行背景匹配,最后对配准后的图像利用帧间差分法提取运动目标.实验表明,该算法具有一定的鲁棒性,能有效地检测动态背景下的运动目标.     

几种开发通用题库管理系统方法的比较

研究了几种通用题库管理系统开发中的试题输入与输出方法,对这几种方法的优缺点进行比较,并给出了对应的程序代码。同时提出了自己开发通用题库管理系统中的试题的输入与输出方法,实验结果表明此方法比其它方法稳定性要好、处理效率要高。     

基于视频图像的绊线检测方法研究

在绊线检测中,高斯模型用于背景建模,背景差分用于前景目标提取,扫描虚拟线上所有点,当检测到目标穿越绊线后,若为双向绊线,系统直接报警;若为单向绊线,利用目标颜色直方图欧拉距离来确定目标运动方向,如果目标运动方向与禁止穿越方向一致,系统报警。实验表明,给出的方法能够准确实现绊线检测。     

视频目标跟踪中的提升分类算法

针对目标跟踪过程中各类图像特征分离背景和目标能力的动态变化,提出一种基于协同训练框架的在线提升分类特征选择算法.该算法采用两组特征描述目标与背景区域各像素,并分别训练一在线提升分类器对特征组进行选择,然后综合分类结果,得到最优似然图像,基于该似然图像,采用粒子滤波对目标进行琅踪并通过图像处理方法获得最佳前景分割图.该方法的主要优点是仅需对首帧图像进行训练,并在跟踪过程中通过协同训练在线更新提升分类器.同时,实验表明该算法运算速度快,并能适应环境光照变化、遮挡等恶劣条件.     

基于嵌入式系统的图像对象分割

随着计算机科学和微电子技术的迅速发展,以计算机技术、集成电路技术和软件技术为核心的嵌入式系统的信息处理能力迅速提升,为基于嵌入式系统的图像信息处理提供了硬件基础。本文以ARM920T结构的S3C2410处理器为核心,构建了一个嵌入式图像处理平台,在嵌入式Linux操作系统的支撑下,实现了嵌入式图像的采集、图像灰度化和基于迭代法的图像对象分割,实验证明本系统是切实可行,为进一步的研究提供了基础。     

A scalable constraint-based Q-hash indexing for moving objects

In this paper, we develop a Q-hash index structure to efficiently store the position of moving objects. An environment of moving objects contains continuously changing locations which are hard to index using traditional index structures such as R-trees, QuadTrees and their variants. In order to answer the queries accurately, one of the problems faced in storing these positions is the number of updates that have to be made to the database whenever locations change. The high maintenance overhead on updates leads to performance degradation of these index structures; additionally, it makes the database very bulky which results in very poor performance in terms of query execution time. One of the main objectives of the structure we propose is to minimize the number of updates to the database to an optimal number so that the accuracy and response time of the query result are not compromised and at the same time the number of wireless communications can be reduced. The indexing is done using a hashing technique where the hashing function makes use of a region based QuadTree structure. To improve the efficiency of the query processing our index structure helps us define constraints over speed, direction and location of the moving object at the device level which controls the number of updates. In addition, in order to answer different query types efficiently at all levels we propose a three-tier (moving object, regional server, central repository) architecture. Our extensive performance evaluation and comparison of the proposed technique concludes that our scheme outperforms existing Q + R-tree and QuadTree in terms of range query execution time by a high order of magnitude.