首页 | 本学科首页   官方微博 | 高级检索  
相似文献
 共查询到20条相似文献,搜索用时 140 毫秒
1.
针对高分辨率一维距离像(HRRP)多极化特征信息融合目标识别带来的数据量剧增问题,提出一种基于低维平移不变特征向量和多分类器动态组合的识别方法。该方法首先提取单极化HRRP序列的3种一维特征组成平移不变的特征向量,然后通过动态组合的方法生成总分类器组合进行分类,最后采用加权投票算法融合4种单极化HRRP的分类结果。实验结果显示,该方法在缩减数据规模的同时,有效利用极化信息,得到了较高的分类正确率。  相似文献   

2.
车型识别是智能交通系统中的一个重要组成部分.利用计算机视觉、模式识别等理论方法,对车型的有效特征提取与分类识别进行了深入的研究,提出了一种多特征融合的模糊聚类车型识别方法.通过自动调整各维特征的加权系数来调整特征对分类的贡献,结果表明达到了较好的分类效果.  相似文献   

3.
该文提出了一种利用不变矩和支持向量机(SVM)对图像进行识别的方法。该方法提取图像的7个不变矩作为特征矢量,对得到的特征量应用支持向量机进行图像分类和识别。该文通过试验验证了此方法的有效性。  相似文献   

4.
特征点和不变矩结合的遥感图像飞机目标识别   总被引:2,自引:0,他引:2       下载免费PDF全文
传统的飞机目标识别算法一般是通过目标分割,然后提取不变特征进行训练来完成目标的识别。但是,对于实际情况比较复杂的遥感图像飞机目标,至今没有一种适合多种机型的分割和识别算法。针对现有识别算法的不足,本研究提出一种基于特征点空间分布、颜色不变矩和Zernike不变矩相结合的遥感图像飞机目标识别算法。方法:首先,对预处理后的遥感图像和模板图像进行小波变换,在低分辨率图像下采用圆投影特征进行粗匹配,确定候选目标;粗匹配结束后,提取高分辨率图像的多尺度Harris-laplace角点,并画出Delaunay三角网,同时提取出颜色不变矩和Zernike不变矩;然后使用欧氏距离作为这三种特征的相似性度量,并和样本图像进行加权匹配;最后选取欧式距离最小的图像作为最终的识别目标。结果:实验表明,本文算法飞机检测精度比现有算法高2.2%,飞机识别精度比现有算法高1.4%-10.4%。该算法能从遥感图像中精确识别出十大飞机目标,并对背景、噪声、视角变化等多种干扰具有良好的鲁棒性。结论:提出了一种基于特征点空间分布、颜色不变矩和Zernike不变矩相结合的飞机识别算法,该算法使用了图像的多种信息,包括特征点和不变矩,有效地克服了使用单一特征无法描述多种信息的不足。实验结果表明,本文采用基于特征点和不变矩的飞机识别算法比其他算法具有更强的抗干扰能力和识别精度。  相似文献   

5.
一种有效的SAR图像目标识别方法   总被引:1,自引:0,他引:1       下载免费PDF全文
SAR图像目标识别是SAR图像应用中非常重要的环节,但由于SAR图像中相干斑噪声的存在,使得传统方法不能很好地对SAR图像进行分类识别。结合不变矩特征提取和支持向量机分类方法的优势,提出了一种有效的SAR图像目标识别方法,采用该方法对含有飞机和坦克目标的SAR图像进行了目标识别实验,取得了较好的识别效果。  相似文献   

6.
为了识别退化的交通标志图像,提出了一种新的分类算法。该算法在处理图像的退化问题时,采用模糊—仿射不变距直接提取图像的特征而不需要图像的清晰化处理;在利用模糊—仿射不变距提取图像特征的基础上,采用递归正交最小二乘算法设计了一种新的径向基概率神经网络分类器。仿真结果表明:模糊—仿射不变距是一种有效的处理退化的交通标志图像的方法,所设计的径向基概率神经网络分类器不仅具有精简的结构,而且,具有较好分类和推广性能。  相似文献   

7.
文章提出了一种基于模糊相似测量的小类别数多字体汉字及数字识别方法.该方法通过模糊逻辑处理,直接将字符的二值化图像转换成基于非线性加权相似函数的模糊样板,然后通过分类模糊模型的统计,相似性测量样板的分级组合和基于规则的分类进行识别.实验表明,该方法用于小类别数多字体汉字及数字识别的效果良好.  相似文献   

8.
基于机器视觉和神经网络的烧结质量预测   总被引:1,自引:0,他引:1       下载免费PDF全文
应用机器视觉和人工神经网络理论提出了对烧结质量在线判断的一种模式识别方法。以某烧结厂为研究背景,分析影响烧结质量的视觉特征,从烧结机机尾摄取断面图像并进行处理,用图像的空间低阶矩描述目标的视觉特征,从而可以选出对分类识别最有效的特征作为人工神经网络的输入,构造改进的BP神经网络分类器,实现在线判断烧结质量,实验证明了该方法有效可行。  相似文献   

9.
邹承明  罗莹  徐晓龙 《计算机应用》2018,38(7):1853-1856
针对单一特征表示的局限性会导致细粒度图像分类准确度不高的问题,提出了一种基于卷积神经网络(CNN)和尺度不变特征转换(SIFT)的多特征组合表示方法,综合考虑对目标整体、关键部位和关键点的特征提取。首先,分别以细粒度图像库中的目标整体和头部区域训练CNN得到两个网络模型,用来提取目标的整体和头部CNN特征;然后,对图像库中所有目标区域提取SIFT关键点并通过K均值(K-means)聚类生成码本,再将每个目标区域的SIFT描述子通过局部特征聚合描述符(VLAD)参照码本编码为特征向量;最后,组合多种特征作为最终的特征表示,采用支持向量机(SVM)对细粒度图像进行分类。使用该方法在CUB-200-2011数据库上进行实验,并与单一的特征表示方法进行了比较。实验结果表明,该方法与基于单一CNN特征的细粒度图像分类相比提升了13.31%的准确度,证明了多特征组合对细粒度图像分类的积极作用。  相似文献   

10.
基于改进型LBP算法的植物叶片图像识别研究   总被引:1,自引:0,他引:1       下载免费PDF全文
为了解决LBP算法抽取的纹理特征仅考虑了邻域像素的特征,忽略关键的局部和全局特征的问题,提出一种基于改进型LBP算法的WCM-LBP植物叶片图像特征提取方法。该算法融合了加权局部均值算法WRM-LBP和加权全局均值算法WOM-LBP,通过提取叶片基于区域的关键几何特征和纹理特征对LBP特征描述符进行加权改造,并采用加权局部均值和加权全局均值代替传统的中心像素点,最后将叶片图像的R、G和B通道颜色分量和灰度值作为特征输入矩阵进行图像分析。该算法结合特征加权的模糊半监督聚类算法(SFFD)应用于经典的Flavia、Swedish、Foliage以及自测图片集等4种植物叶片图像数据集中进行实验。实验结果表明,该算法具有很强的鲁棒性,能够有效区分机器视觉下植物叶片图像的关键性识别特征,有效解决叶片图像的分类识别中关键特征的描述问题。  相似文献   

11.
A new method of selection and reduction of system feature in pattern recognition based on rough sets is proposed in this paper. Its basic idea is that the classification ability of system feature is evaluated with the classification affection introduced by removing some redundant feature or unimportant feature in the system and comparing the final classification result to find out useful feature in pattern recognition. Eventually, reduction and optimum combination of feature sets are performed by this method. An example of selection and reduction of system feature in pattern recognition based on rough sets shows the correctness and effectiveness of the method.  相似文献   

12.
Feature Extraction Using Independent Components of Each Category   总被引:1,自引:0,他引:1  
We describe an application of independent component analysis (ICA) to pattern recognition in order to evaluate the effectiveness of features extracted by ICA. We propose a recognition method suitable for independent components that consists of modules for each category. A module has two parts: feature extraction and classification. Features are independent components estimated by ICA and outputs of modules are candidates for categories. These candidates are combined and categories are decided with a majority rule. This recognition method is applied to two tasks: hand-written digits in the MNIST database and acoustic diagnosis for a compressor as real-world tasks. A FastICA algorithm is applied to extracting independent features in the proposed method. Through recognition experiments, we demonstrate that the ICA of each category extracts useful features for these tasks and the independent components are superior to the principal components in recognition accuracy. Manabu Kotani - Deceased  相似文献   

13.
基于条形码的结帐系统存在操作繁琐等一些问题,为了解决零售业结账服务中排队的难题,提出一种以SURF(Speeded Up Robust Features)特征匹配为主,主颜色特征数字编码分类、形状特征数字编码分类为辅的商品快速识别算法。基于特征的方法具有压缩信息量、精度高等优点,成为目前研究的热点。但是传统图像特征识别算法,存在特征维度多、计算量大、运行速度慢等缺点,限制了其应用。本文将其中的主颜色和形状特征进行数字编码分类,之后利用高效识别算法SURF进行准确识别,很好地克服了以上缺点。实验表明,本文算法运行速度快,识别性能好,为零售业的结账服务提供了便利。  相似文献   

14.
15.
Automatic solder joint inspection   总被引:3,自引:0,他引:3  
The task of automating the visual inspection of pin-in-hole solder joints is addressed. Two approaches are explored: statistical pattern recognition and expert systems. An objective dimensionality-reduction method is used to enhance the performance of traditional statistical pattern recognition approaches by decorrelating feature data, generating feature weights, and reducing run-time computations. The expert system uses features in a manner more analogous to the visual clues that a human inspector would rely on for classification. Rules using these cues are developed, and a voting scheme is implemented to accumulate classification evidence incrementally. Both methods compared favorably with human inspector performance  相似文献   

16.
基于典型相关分析的组合特征抽取及脸像鉴别   总被引:14,自引:0,他引:14  
利用典型相关分析的思想,提出了一种基于特征级融合的组合特征抽取新方法.首先,抽取同一模式的两组特征矢量,给出描述两组特征矢量之间相关性的判据准则函数;然后依此准则抽取它们的典型相关特征,构成有效鉴别特征矢量用于识别.该方法巧妙地将两组特征矢量之间的相关性特征作为有效判别信息,既达到了信息融合之目的,又消除了特征之间的信息冗余,为两组特征融合用于分类识别提供了新的思路.此外,从理论上进一步剖析了所提出的方法之所以能有效地用于识别的内在本质.在Yale和ORL标准人脸数据库上的实验结果证实了所提算法的有效性和稳定性,而且识别率大大高于用单一特征进行识别的结果.  相似文献   

17.
A new approach to time-frequency transform and pattern recognition of non-stationary power signals is presented in this paper. In the proposed work visual localization, detection and classification of non-stationary power signals are achieved using hyperbolic S-transform known as HS-transform and automatic pattern recognition is carried out using GA based Fuzzy C-means algorithm. Time-frequency analysis and feature extraction from the non-stationary power signals are done by HS-transform. Various non-stationary power signal waveforms are processed through HS-transform with hyperbolic window to generate time-frequency contours for extracting relevant features for pattern classification. The extracted features are clustered using Fuzzy C-means algorithm and finally the algorithm is optimized using genetic algorithm to refine the cluster centers. The average classification accuracy of the disturbances is 93.25% and 95.75% using Fuzzy C-means and genetic based Fuzzy C-means algorithm, respectively.  相似文献   

18.
一种有效的手写体汉字组合特征的抽取与识别算法   总被引:2,自引:0,他引:2  
基于特征融合的思想,从有利于模式分类的角度,推广了典型相关分析的理论,建立了广义的典型相关分析用于图像识别的理论框架。在该框架下,首先利用广义的典型相关判据准则函数,求取两组特征矢量的广义投影矢量集,构成一对变换矩阵;然后根据所提出的新的特征融合策略,对两种手写体汉字特征进行融合,所抽取的模式的相关特征矩阵,在普通分类器下取得了良好的分类效果,优于已有的特征融合方法及基于单一特征的PCA 方法和FLDA 方法。  相似文献   

19.
多字体印刷藏文字符识别   总被引:5,自引:1,他引:5  
藏文字符识别系统是中文多文种信息处理系统的重要组成部分,但至今国内外的研究基本处于空白。本文提出了一种基于统计模式识别的多字体印刷藏文字符识别方法:从字符轮廓中抽取方向线素特征,利用线性鉴别分析(LDA)压缩降维后得到紧凑的字符特征向量。采用基于置信度分析的两级分类策略,设计了带偏差欧氏距离分类器(EDD)完成高效的粗分类,细分类采用修正二次鉴别函数(MQDF)。通过实验选取恰当的分类器参数后,在容量为177,600字符(300样本/字符类)的测试集上的识别率达到99.79%,证明了该方法的有效性。  相似文献   

20.
Damage detection in structures is one of the research topics that have received growing interest in research communities. While a number of damage detection and localization methods have been proposed, very few attempts have been made to explore the structure damage classification problem. This paper presents an Artificial Immune Pattern Recognition (AIPR) approach for the damage classification in structures. An AIPR-based structure damage classifier has been developed, which incorporates several novel characteristics of the natural immune system. The structure damage pattern recognition is achieved through mimicking immune recognition mechanisms that possess features such as adaptation, evolution, and immune learning. The damage patterns are represented by feature vectors that are extracted from the structure’s dynamic response measurements. The training process is designed based on the clonal selection principle in the immune system. The selective and adaptive features of the clonal selection algorithm allow the classifier to evolve its pattern recognition antibodies towards the goal of matching the training data. In addition, the immune learning algorithm can learn and remember different data patterns by generating a set of memory cells that contains representative feature vectors for each class (pattern). The performance of the presented structure damage classifier has been validated using a benchmark structure proposed by the IASC–ASCE (International Association for Structural Control–American Society of Civil Engineers) Structural Health Monitoring (SHM) Task Group and a three-story frame provided by Los Alamos National Laboratory. The validation results show that the AIPR-based pattern recognition is suitable for structure damage classification. The presented research establishes a fundamental basis for the application of the AIPR concepts in the structure damage classification.  相似文献   

设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号