一种道路车辆监控视频中的关键帧提取方法

针对道路监控视频中特定车辆图像序列的关键帧提取问题,在运动对象检测的基础上,提出一种关键帧提取方法。将积分通道特征和面积特征作为图像特征描述子,结合Ada Boost训练分类器,实现道路监控视频车辆序列图像中关键帧的提取。通过运动对象前景检测技术获得出现在监控区域的运动车辆最小外接矩形图像序列,选择满足监控分析需求(车牌清晰度高,能判断车型)的若干帧作为正样本,其他不满足监控分析需求的作为负样本,提取样本图像的面积特征和积分通道特征,利用Ada Boost方法训练得到一个分类器,使用Ada Boost分类器对测试样本进行分类,根据打分规则提取关键帧。实验结果表明,该方法能提取运动车辆从进入到离开监控区域的序列图像帧中最清晰的图像,实现道路车辆监控视频分析数据的有效压缩。     

Velocity measurements and flow field characteristic analyses in a turbo air classifier

Turbo air classifier is one of the most widely used powder classification equipment. The rotor cage as a rotary component can create a forced centrifugal field, so it is a key part for turbo air classifier. In order to investigate the effect of structural variations of the rotor cage on flow field characteristics, three dimensional velocity measurements of the annular region in a turbo air classifier equipped with two different rotor cage bottom plates (A and B type) are performed by laser Doppler velocimeter (LDV). It is found that the different bottom plates have different axial and tangential velocity distributions in the annular region. However, the structural variations of the rotor cage have hardly any effect on the radial velocity. Based on the classification principle, the relation between the classification performance and the flow field characteristics is investigated in great detail.The results of the flow field measurements were tested by the classification experiments carried out with cement raw meal and ground calcium carbonate. The results demonstrate that B type bottom plate can realize the production of narrow particle size distributions, so it is more favorable for classification than A type bottom plate. Classification experiment results are in good agreement with the results of the flow field measurements.     

基于SURF与Hough森林的人脸检测研究

为实现复杂场景中的人脸检测与定位,提出了一种基于快速鲁棒特征SURF与Hough森林的人脸检测算法。采用SURF局部特征构建Hough森林分类器,每个叶子节点存储类别信息与特征点到达目标中心的偏移量,在图像局部外观与Hough投票之间建立映射关系,生成有监督的判别式的码本,获得可靠的概率Hough投票,以此预测目标中心位置,提高了检测精度。与此同时,采用SURF局部特征提取图像兴趣点有助于减小计算量、加快检测速度。实验证明了所提算法的有效性。     

基于条件对数似然的BP神经网络多类分类器

BP神经网络分类器存在收敛速度慢的缺陷,为了提高分类器性能,针对这一缺陷对BP算法进行改进. 提出将条件对数似然（CLL）准则融入到监督性BP神经网络多类型分类过程中,利用CLL的可分解性优势,计算测试样本的条件概率,在误差反向传播时利用条件概率对权值进行相应的加权降权操作,简化误差反馈过程中的计算量. 在实验中对改进算法的收敛速度和准确率进行了测试,说明了该算法的有效性及实用性.     

贝叶斯网络分类器近似学习算法

贝叶斯网络在很多领域应用广泛,作为分类器更是一种有效的常用分类方法,它有着很高复杂度,这使得贝叶斯网络分类器在应用中受到诸多限制。通过对贝叶斯网络分类器算法的近似处理,可以有效减少计算量,并且得到令人满意的分类准确率。通过分析一种将判别式算法变为产生式算法的近似方法,介绍了这种算法的近似过程,并将其应用在了贝叶斯网分类算法中。接着对该算法进行分析,利用该算法的稳定性特点,提出Bagging-aCLL 集成分类算法,它进一步提高了该近似算法的分类精度。最后通过实验确定了该算法在分类准确率上确有不错的表现。     

基于优化支持向量机的人脸表情分类

分析了支持向量机（support vector machine, SVM）目前主要存在的问题和参数选择对分类性能的影响后, 提出了以改进粒子群算法优化SVM关键参数的优化SVM算法。将加入拥挤度因子的微粒群算法引入到SVM中, 在不牺牲泛化性能的前提下, 对其参数进行优化, 增加了SVM初始化参数的多样性, 减慢了局部搜索, 促进其在全局范围内的寻优搜索, 以有效克服SVM算法过分依赖初始值和容易陷入局部极小值的缺点, 并利用由粗到精的策略构造多层SVM人脸表情分类器, 在提高准确率的基础上加快分类的速度。实验证明, 新算法具有速度快、准确率高的优点。     

Gasoline quality prediction using gas chromatography and FTIR spectroscopy: An artificial intelligence approach

This paper reports on analysis of 45 gasoline samples with different qualities, namely, octane number and chemical composition. Measurements of data from gas chromatography and IR (FTIR) spectroscopy are used to gasoline quality prediction and classification. The data were processed using principal component analysis (PCA) and fuzzy C means (FCM) algorithm. The data were then analyzed following the neural network paradigms, hybrid neural network and support vector machines (SVM) classifier. The IR spectra were compressed and de-noised by the discrete wavelet analysis. Using the hybrid neural network and multi linear regression method (MLRM), excellent correlation between chemical composition of the gasoline samples and predicted value of the octane number was obtained. About 100% correct classification for six different categories of the gasoline was achieved, each of which has different qualities.     

Measurement of soybean fatty acids by near-infrared spectroscopy: Linear and nonlinear calibration methods

A key element of successful development of new soybean cultivars is availability of inexpensive and rapid methods for measurement of FA in seeds. Published research demonstrated applicability of NIR spectroscopy for FA profiling in oilseeds. The objectives of this study were to investigate the applicability of NIR spectroscopy for measurement of FA in whole soybeans and compare performance of calibration methods. Equations were developed using partial least squares (PLS), artificial neural networks (ANN), and support vector machines (SVM) regression methods. Validation results demonstrated that (i) equations for total saturates had the highest predictive ability (r ²=0.91–0.94) and were usable for quality assurance applications, (ii) palmitic acid models (r ²=0.80–0.84) were usable for certain research applications, and (iii) equations for stearic (r ²=0.49–0.68), oleic (r ²=0.76–0.81), linoleic (r ²=0.73–0.76), and linolenic (r ²=0.67–0.74) acids could be used for sample screening. The SVM models produced significantly more accurate predictions than those developed with PLS. ANN calibrations were not different from the other two methods. Reduction in the number of calibration samples reduced predictive ability of all equations. The rate of performance degradation of SVM models with sample reduction was the lowest.     

一种改进的朴素贝叶斯分类器在文本分类中的应用研究

文本分类是数据挖掘领域中重要的研究分支.通过对自适应遗传算法和朴素贝叶斯分类器的研究,提出一种基于自适应遗传算法的朴素贝叶斯分类算法.将该算法应用于中文文本分类中,可以生成最优贝叶斯分类器及最优属性集合,提高分类精度.     

基于KSVD和PCA的SAR图像目标特征提取

提出一种基于核的奇异值分解(KSVD)与主成分分析(PCA)相结合的SAR图像目标的组合特征提取方法。该方法首先利用核的奇异值分解得到图像非线性的代数特征,然后进一步经过PCA变换得到图像的最终分类特征。实验中,将本文提出的KSVD+PCA两步特征提取方法与PCA、SVD、KPCA、KSVD方法分别结合简单、快速的最近邻分类器在MSTAR坦克数据上进行了比较,实验结果表明,KSVD+PCA方法不仅有效地提高了目标的正确识别率,而且大大降低了对目标方位的敏感度,在目标方位信息未知的情况下,识别率可达到95.75%,是一种有效的SAR图像目标特征提取方法。