Support vector regression based color image restoration in YUV color space

A support vector regression(SVR) based color image restoration algorithm is proposed.The test color images are firstly mapped into the YUV color space,and then SVR is applied to build up a theoretical model between the degraded images and the original one.Performance comparisons of the proposed algorithm versus traditional filtering algorithms are given.Experimental results show that the proposed algorithm has better performance than traditional filtering algorithms and has less computation time than iterat...     

驾驶人手机通话行为中基于图像特征决策融合的手势识别方法

为鲁棒检测自然环境中驾驶人的通话行为, 提出了一种驾驶人手机通话手势的识别方法。运用Adaboost算法检测驾驶人面部区域, 在YC_gC_r色彩空间中分别对面部肤色亮度分量和色度分量进行稀疏网格间隔采样, 由此建立了肤色的高斯分布模型; 针对驾驶室光照强度的不均匀性, 提出了肤色分量的漂移补偿算法, 建立了适应光照变化的在线肤色模型, 以准确分割左右手部肤色区域; 运用HOG算法获取手部肤色区域的2 376维HOG特征向量, 运用PCA方法将HOG特征降至400维; 同时提取手部肤色区域的PZMs特征, 并采用Relief算法筛选出权重最大的8个PZMs特征向量, 建立了融合PCA-HOG特征和Relief-PZMs特征的通话手势支持向量机分类决策。试验结果表明: 基于PCA-HOG特征的手势识别率为93.1%, 对光照变化的鲁棒性较好, 但易受到手部与头部转动的干扰; 基于Relief-PZMs特征的手势识别率为91.9%, 对于头部与手部姿态的耐受度较好, 但光照鲁棒性较差; 基于PCA-HOG和Relief-PZMs多元特征融合方法的手势识别率达到94.5%, 对光照波动、手部与头部转动等干扰条件具有较好的适应性。      

���ڳ���ǰ��������Ĺ��������

轨道交通线路净空安全是确保列车平稳、不间断运行的基础.由于轨间异物对行车安全产生严重影响,所以基于非轨道电路的轨道异物入侵检测系统在铁路系统中具有十分广泛的应用前景.本文提出了一种基于移动车载摄像机检测轨间异物的方法.首先,通过钢轨识别算法自动定位钢轨位置,并确定列车前方轨道是否有其它列车或公路车辆等大型异物,若有则进行报警;之后,基于边缘检测的异物检测算法自动检测轨间可疑小异物,同时提取可疑异物的尺度信息和颜色索引参数等相关特征;最后,用支持向量机（SVM）来对可疑小异物区域进行分类和辨识.车载实验结果表明,该方法可以有效地检测出轨间异物.     

Human and bicycle detection system using range sensor

This paper presents a full-scale solution to the detection of the traffic data using laser device. Range images, gathered by a particular laser camera, are used in the multi-threshold segmentation. The multi-threshold segmentation is based on the height of the moving objects. In order to get the precise height of the moving objects, mapping of the original terrain is performed on the first step. On each layer, the clustering algorithm called iteration-self organizing data analysis techniques algorithm (ISODATA) is conducted afterwards. Kalman filtering technique is applied to recognize and track the moving objects. Extensive experiments show that these algorithms are effective in object recognition and tracking, as well as robust in the applications.     

基于多SVM分类器融合的高速公路异常事件检测方法

提出了一种利用多SVM分类器对高速公路中的复杂交通信息进行有效融合的异常事件检测方法.首先,将初始训练集划分为互不重叠的子集,为每个子集训练分类器.给定一个输入向量,利用分类器求得其所属的类别标签,并计算出该向量对特定簇的隶属度.其次,利用概率方法将多SVM分类器分类结果进行融合,得到最终分类结果.接下来,将“车流量”、“行车速度”、“道路占用率”、“相邻监测站的车流量差值”、“速度差值”以及“道路占用率差值”等交通参数表示为特征向量,分别输入到经过训练的SVM分类器,并将多SVM分类器融合后的分类结果作为判别异常事件的依据.最后,从5个具有代表性的高速公路路段采集到的交通数据构造实验数据集.实验结果表明,对比单一SVM和LS-SVM,文章提出的基于多SVM分类器融合的高速公路异常事件检测方法可以有效提高高速公路异常事件检测的准确性和可靠性,弥补了仅使用单一交通参数进行异常事件检测的不足.     

基于特征脸和LS-SVM分类器的人脸性别分类

提出使用特征脸和最小二乘支持向量机(LS-SVM)分类器相结合进行人脸性别分类.我们首先从训练图像中求得特征脸空间,然后将训练集和测试集图像投影到特征脸空间得到投影系数.使用训练样本投影系数训练LS-SVM分类器,对训练图像和测试图像进行分类试验,同时计算出分类准确率,实验结果表明LS-SVM分类要比其他分类算法有更好的优越性.在实验中我们也使用交叉验证来确定特征脸数目和核函数参数.     

单相SAPF系统谐波电压检测与控制的仿真

为了改善用户侧电压质量与谐波问题,通过分析单相串联型有源电力滤波器（SAPF）的结构、原理及特点,提出了单相SAPF系统谐波检测的方法;同时借鉴单相SVPWM技术原理,将开关模式优化的SVPWM技术应用于单相SAPF系统中,提高了系统波形控制精度;利用Matlab软件,建立了基于上述谐波检测和系统波形控制的单相SAPF系统的仿真模型,并利用该模型对动态电压跌落与闪变等电压故障状态进行了动态补偿的仿真.仿真结果显示：本文提出的SAPF滤波系统不但具有良好的滤波性能,而且可以保障负载不受系统电压故障的影响,具有良好的工程应用前景.     

面向多目标跟踪的密集行人群轨迹提取和运动语义感知

针对基于视频监控的密集行人群识别难度大，运动轨迹提取困难，运动语义信息挖掘不足 等问题，本文提出基于多目标跟踪FairMOT框架及K-means聚类的行人轨迹捕获和运动语义信 息感知方法。首先，利用多目标跟踪算法提取视频中行人群目标过街时的运动轨迹特征向量；然 后，通过分析轨迹坐标的数值分布特点，设计了一种协方差滤波算法STCCF，以检测和剔除“准静 态轨迹”，得到行人群目标运动轨迹簇；最后，针对已提取的轨迹簇，应用K-means聚类方法，选取 S系数(Silhouette Coefficient)和DB指数(Davies Bouldin Index)两个指标，感知行人聚集和消散点 等场景语义特征。实验分析表明，算法从提取到的2689条轨迹中辨识出179条“准静态轨迹”，检 出率为81.73%；视频场景中的行人目标源点和消失点的解析结果与人工辨识结果吻合，验证了密 集行人群轨迹提取和运动语义信息感知方法的有效性。本文研究可为数据驱动的行为预测和轨 迹建模提供基础。     

一种新的肤色区域分割提取方法

提出一种新的肤色区域分割提取方法,并详细叙述了该方法的步骤和实现算法.在人脸检测和识别项目中,应用该算法来提取各肤色区域,经大量实验说明,该方法是有效的.     

一种新的图像中人眼定位方法

采用一种人脸-人眼两层结构的定位方法.首先利用YCbCr空间的肤色模型粗定位可能的人脸区域,然后利用支持向量机的训练结果组成分类器,对人眼进行初检,最后根据人脸特征完成信息融合,最终标定人眼.实验结果表明:该方法在简单及复杂背景下都能够得到较高的定位速度和精确度,具有较好的鲁棒性.     

结合区域分割和Wishart分类器的机场跑道区域快速检测方法

提出了一种结合区域分割和Wishart分类器的极化合成孔径雷达图像机场跑道区域快速检测方法; 利用简单线性迭代聚类算法分割极化合成孔径雷达图像, 并将分割得到的超像素作为后续分类处理的基本单元; 采用一种优化后的距离度量方式给超像素分配类别标签, 解决了传统Wishart距离度量因子冗余运算量大的问题; 分析了机场跑道区域像素的极化散射特性, 利用机场跑道区域的弱散射特性从分类结果中提取感兴趣区域; 利用机场跑道的结构特征筛选辨识感兴趣区域, 进而确定机场跑道区域的准确位置; 利用极化合成孔径雷达实测数据测试了算法的有效性, 并与传统基于像素的检测结果进行对比。试验结果表明: 该算法在复杂大场景下能够快速有效检测出机场跑道区域, 检测出的跑道轮廓清晰, 结构比较完整; 采用简单线性迭代聚类算法预处理图像极大地降低了后续处理的复杂性; 针对墨西哥湾试验数据, Wishart分类器处理单元个数分别是Freeman+Wishart算法和FCM+Wishart算法的1.0%和2.4%, 整个检测过程耗时分别为Freeman+Wishart算法和FCM+Wishart算法的9.9%和27.1%;针对大岛试验数据, Wishart分类器处理单元个数分别是Freeman+Wishart算法和FCM+Wishart算法的1.0%和2.6%, 整个检测过程耗时分别为Freeman+Wishart算法和FCM+Wishart算法的14.0%和31.8%。可见, 所提检测方法的实时性能优于基于像素的检测方法。      

用于检测疲劳驾驶的自适应脸部定位方法

针对疲劳驾驶视觉检测技术中驾驶员脸部定位快速性与准确性难以同时满足的情况,提出了一种将肤色和面部几何特征相结合的自适应脸部定位方法。该方法对YCbCr空间肤色相似度和传统Otsu闾值分割算法进行改进,降低了计算复杂度,采用新的区域连通和区域标记算法定位出脸部位置。最后,根据人脸几何特征对脸部区域做进一步确认：试验结果表明,该方法能对驾驶员图像脸部进行快速、准确定位。     

Real-time follow-up head tracking in dynamic complex environments

In the modes of both object motion and camera motion, an enhanced Camshift algorithm, which is based on suppressing similar color features of background and on joint color probability density distribution image, is proposed to real-time track head in dynamic complex environment. The system consists of face detection module, head tracking module and camera control module. When tracking fails, a self-recovery mechanism is introduced. At first the Adaboost face detector based on Haar-like features is implemented to find frontal faces, the false positive is filtered according to the skin color criterion, and the true face is used to initialize the tracking module. In hue saturation value (HSV) colorspace, the hue-saturation (H-S) histogram of face skin and the saturation-value (S-V) histogram of hair are built to produce the joint color probability density distribution image, and this is intended to realize the head tracking with arbitrary pose. During tracking, region of interest (ROI) is introduced, and the color probability density distribution of a specified background area outside the ROI is learned, similar color features in the head are suppressed according to the learning result. The background suppression step is intended to resolve the problem that the tracker maybe fails when the head is distracted by backgrounds having similar colors with the head. A closed loop control model based on speed regulation is applied to drive an active camera to center the head. Once tracking drift or failure is detected, the system stops tracking and returns to the face detection module. Our experimental results show that the presented system is well suitable for tracking head with arbitrary pose in dynamic complex environments, also the active camera can track moving head smoothly and stably. The system is computationally efficient and can run in real-time completely.     

Image fusion scheme based on nonsubsampled contourlet and block-based cosine transform

A kind of nonsubsampled contourlet and block-based cosine transform (NSCBCT) is developed, and its application in image fusion is studied in this paper. The construction of filtering banks is based on the nonsubsampled contourlet transform (NSCT) and block-based discrete cosine transform (B-DCT). We combine NSCT and B-DCT to design filters that lead to NSCBCT with better singularity representation than either of them in isolation. A design framework based on the hybrid approach is proposed, which allows for the fast implementation based on NSCT and B-DCT respectively. In addition, a new image fusion scheme based on NSCBCT for multispectral and panchromatic satellite images is proposed. Firstly, because it adopts NSCT, the fused satellite images have higher spatial resolution than those based on wavelets. Secondly, based on the localized high frequency information provided by B-DCT, the proposed fusion scheme can reduce the spectral distortion of the fused image further. Experimental results show that the proposed fusion method is able to increase the spatial resolution and reduce the spectral distortion of the fused image at the same time.     

An Improvement of Russo's Fuzzy Noise-Filtering for Image Processing

We adapted Russo's image noise filtering technique by introducing an adaptive parameter, noise possibility. This parameter potentially indicates whether a pixel is noise. The improved Russo's method was applied to process the Lena image and a remote sensing image corrupted by noise. The results show that it has a stronger noise filtering ability than other traditional techniques such as Wiener filtering.     

An SNN Ontology Based Environment Monitoring Method for Intelligent Irrigation System

In order to realize high precision of environment parameters detection in irrigation applications, a sensor and sensor network (SSN) ontology based data fusion method is proposed. An SSN sub-ontology for soilstate monitoring is revised, which includes the sensing devices hierarchies and measurement properties selection according to the detection feature interests. As for sensor data processing, a tuning data method by data pool filtering and clustering is adopted, as well as a useful data fusion method for multi-sensor system. The testing results show that both the accuracy and efficiency of the proposed method are higher after related filtering and clustering process, which enables a thorough monitoring for intelligent irrigation systems and can be extended into environment monitoring and control applications.     

基于监控视频的高速公路能见度检测方法及系统

针对我国高速公路能见度天气的全面检测与预警处置的迫切需求,提出了一种基于监控视频的高速公路能见度检测方法.该方法从监控摄像头提取固定距离参照物的图像,根据图像的灰度失真状况计算能见度值,并通过数据融合的手段对多监控摄像头计算结果进行综合;并设计和开发了一款能见度监测与预警处置信息化系统,利用获取的实时能见度数据,从局部现场和宏观路网两个方面支撑能见度预警、应急处置、交通流调控、信息发布等智能交通业务.实验分析结果表明,该能见度计算方法的平均检测误差为13.4%,能够满足我国高速公路能见度检测的精度和计算性能要求;本文方法及系统可以充分复用高速公路沿线广泛架设的视频监控机电系统,可应用于我国高速公路的运营管理及效率提升等工作中.      

基于Co—Training策略的视频广告文本检测

同普通视频节目相比,视频广告中的文本具有更为复杂的表现形式．为实现这类文本有效的定位,通过将文本检测视为一种特殊纹理的分类问题,提出一种基于改进的Co—training策略的视频广告文本检测方法,采用两种相对独立的纹理描述子,从多视角来强化文本特性描述．另外,针对Co-training协同学习机制中容易引入噪声样本的问题,提出了一种改进的结合Bootstrap思想的Co—training算法,在两个相对独立的特征空间中交互选择典型样本,以达到提高分类器泛化能力的目的．通过实验,本方法在自建的数据库上获得的正确率与查全率相对于其他方法有10％左右的提高．     

基于图像表面梯度的翻拍检测

经过翻拍后的篡改图像能够轻易绕过现有的图像篡改检测系统,这对图像真实性的检测构成严重的安全威胁.本文基于翻拍过程中两次使用相机进行拍摄而引入的非线性响应,使用Lib-SVM设计图像分类器判别翻拍图像和真实图像.翻拍后图像表面梯度值与真实图像相比会产生非线性变化,这使翻拍图像表面梯度值产生异常.本文基于图像表面梯度特性提取相关特征值,使用支持向量机分类器进行翻拍图像和真实图像的判别.实验结果表明,本文提出的特征分类效果良好,可以正确检测翻拍图像.