Classification performance comprehensive evaluation of an air classifier based on fuzzy analytic hierarchy process

The classification performance evaluation goal for an air classifier is usually limited to one of the classification performance indices including cut size, classification precision, Newton classification efficiency and degree of dispersion. This method hardly evaluates these performance indices of an air classifier comprehensively and suitably. In order to evaluate the classification performance truly and synthetically, Fuzzy Analytic Hierarchy Process is used to calculate the weights of the classification performance indices after determining the hierarchical model in the present paper. The dimensionless transformation eliminates the effect of the different dimensions. Then, the comprehensive evaluation value of the classification performance for each experiment is obtained using the linear weighted method. The maximum value corresponds to the best classification performance among these evaluation values. In the present study, a turbo air classifier is used as the classification system and talc powders are used as materials. The best classification performance indices are a cut size of 16.5 μm, a classification precision of 0.59, a Newton classification efficiency of 57%, and a degree of dispersion of 2.13. The corresponding optimal operational parameter combinations are: the feeding speed is 40 kg·h^–1, the air inlet velocity is 5 m·s^–1 and the rotor cage's rotary speed is 1200·min^–1. This assessment method avoids the limitation of evaluating a single classification performance index and the incomplete information derived from single factor experiments. Furthermore, the method also provides quantitative evaluation criteria for the classification performance of an air classifier. In the proposed method, the classification performance indices can be selected and the precedence relation matrix of Fuzzy Analytic Hierarchy Process can be set flexibly according to production requirements.     

一种基于多分类器融合的双色红外成像目标识别方法

针对双色红外成像系统中的自动目标识别问题,提出了一种基于多分类器融合的红外目标识别方法.该方法首先提取目标的形状特征和面貌特征,并设计多个基于不同特征的分类器对目标进行分类;然后对各个分类器的目标分类结果进行决策级融合处理,并采用所提出的决策规则对多分类器融合分类结果进行处理得到最终的目标识别结果.该方法充分利用了目标在多传感器图像中的多种分类特征信息,提高了系统的目标识别效率和精确性.实验结果证实了该方法的有效性.     

基于局部线性变换的HUGO图像隐写检测

Pevny等人2010年提出一种最先进的高度不可检测的自适应空域隐写算法HUGO(Highly Undetectable Steganography),其保护了相邻四个像素的一阶统计特性,具有很强的抗检测性能,目前国内外针对HUGO检测缺乏有效的方法。由于用HUGO隐写算法进行隐写嵌入时改变部分纹理特征,本文提出一种新的检测方法,利用局部线性变换得到纹理残差图像,计算共生矩阵得到22130维特征向量,最后使用集成分类器进行分类。实验结果表明:在嵌入率为0.4bpp时,针对BOSSRank图像集,获得平均82.71%的检测率,优于Hugobreaker的80.3%和Guel的76.8%,并在低于0.4bpp嵌入率时,其检测效果有所提高。     

基于DSP的人工路标识别算法的实现

针对以人工路标作为引导标识的视觉导航技术,提出了一种基于导引灯的人工路标识别算法。首先利用基于灰度信息的识别算法对路标上导引灯进行识别与定位;然后,利用交比不变性进行数字的初步定位;最后用投影法进行数字分割,主成分分析法(PCA)和最小距离分类器的模式识别方法对数字进行识别。算法在ICETEK-DM6437-B评估板上实现。实验表明,该算法具有很高的稳定性并能较好地满足实时性的要求。     

梯度向量直方图的行人检测

行人是道路交通的参与者也是受害者,计算机视觉的行人检测,由于其在车辆辅助驾驶系统中的重要应用价值,成为当前计算机视觉和智能车辆领域最为活跃的研究课题之一。为了快速准确地对行人进行检测,提出了一种梯度向量直方图的行人检测方法。该方法使用高斯平滑进行图像预处理,然后利用直方图相似性,选择一些具有强分辨能力的特征来表征行人,最后用(Haar)特征分类器,从目标和背景中获得的经过筛选的特征。实验结果表明,该方法对行人检测是快速且有效的。     

Residence time distribution of particles in a continuous liquid-solid classifier

Experiments were carried out in a continuous classification column of diameter 191 mm and height 1.47 m with water as the liquid and glass beads as the particles. Initial experiments were carried out with narrowly sized particles and somewhat larger tracer particles which could be separated by screening. The main experiments involved a wide distribution of particle sizes, with injected pulses of tracer particles of slightly higher density separated from samples using an organic liquid of intermediate density. The results show that the mixing behaviour is intermediate between plug flow and perfect mixing. Increasing the discharge rate of underflow product stream decreased both the mean residence time and the degree of classification significantly. At relatively low voidage, small particles settled more quickly and large particles more slowly than predicted on the basis of the approach proposed by Lockett and Al-Habbooby (Trans. Inst. Chem. Eng. 51(1973) 281-292) where each particle fraction is assumed to travel as if it were alone in a suspension of the overall solids concentration.     

Water based fractionation of particles

A feed of particles covering a wide range of particle sizes and densities was placed into a novel separator, referred to as the Reflux Classifier, and fractionated at a series of fluidization velocities. Double fractionation was then achieved when each product was sieved into narrow size fractions. For the coal particles used in this study, the ash % was determined for each of the samples. The ash % correlated strongly with particle density, and hence was a useful proxy for the particle density. A new Fractionation Algorithm was developed, which provided a method for estimating the feed size and density distribution from the Double Fractionation data. The cumulative yield % versus the cumulative ash % curves for each size fraction, and for the overall feed were then determined. The findings from the study were validated against accurate results obtained using the sink-float method. Excellent agreement with the curves obtained using the sink-float method was obtained. The sink-float method requires the use of heavy liquids; these are invariably toxic and expensive, hence a viable alternative is needed. The water based fractionation method, in conjunction with the Fractionation Algorithm, may prove to be a suitable substitute for the sink-float method. Precise details of the method and algorithm are therefore provided in order to permit the adoption of this new approach by other laboratories.     

Discriminatively regularized least-squares classification

Over the past decades, regularization theory is widely applied in various areas of machine learning to derive a large family of novel algorithms. Traditionally, regularization focuses on smoothing only, and does not fully utilize the underlying discriminative knowledge which is vital for classification. In this paper, we propose a novel regularization algorithm in the least-squares sense, called discriminatively regularized least-squares classification (DRLSC) method, which is specifically designed for classification. Inspired by several new geometrically motivated methods, DRLSC directly embeds the discriminative information as well as the local geometry of the samples into the regularization term so that it can explore as much underlying knowledge inside the samples as possible and aim to maximize the margins between the samples of different classes in each local area. Furthermore, by embedding equality type constraints in the formulation, the solutions of DRLSC can follow from solving a set of linear equations and the framework naturally contains multi-class problems. Experiments on both toy and real world problems demonstrate that DRLSC is often superior in classification performance to the classical regularization algorithms, including regularization networks, support vector machines and some of the recent studied manifold regularization techniques.     

Single-stage recovery and concentration of mineral sands using a REFLUX™ Classifier

The REFLUX™ Classifier is a gravity separation device that consists of a system of inclined channels located above a fluidized bed. Here we report for the first time the results obtained processing a minerals sands feed using narrow 6 mm channels that promote a laminar-shear separation mechanism that enhances the separation based on density. The feed had a head grade of approximately 5 wt% heavy minerals of density greater than 2800 kg/m³, with the majority in the size range from 50 to 150 μm. The overall recovery of the heavy minerals component was approximately constant at 85% for solids throughputs of 10–18.5 t/(m² h). The unrecovered heavy mineral had an average density close to that of the gangue, hence was deemed unrecoverable by gravity separation. The recovery of the denser and more liberated zirconium mineral exceeded 95%. So this work demonstrated the potential for the heavy mineral to be upgraded by a factor of 16 or more in a single stage operation.     

选粉机气流压损分析与结构优化

根据选粉机构造与流形将选粉机的流体腔划分为6个子腔体,并采用数值模拟的方法得到各腔体的压损比例,发现送料筒腔体、出口腔体与内腔体这三个腔体的压损之和达到了整机压损的86.66%。针对送料筒腔体的气流对冲损耗,设计了侧环面风口结构。针对内腔体内的旋流损耗和出口腔体内转向损耗,设计了涡流打散锥结构。实验结果表明,所设计的优化结构不影响选粉机的分级性能,且能有效的降低选粉机气流压损,节能比例在14.30%~19.80%之间,节能效果良好。