Description: 本文完成了对唇动身份识别技术几个基本问题的理论研究,并对整个系统加以实现.作为本文研究的实验基础,我们建立了唇动方式身份识别数据库(HITLUDB),
该库目前包含30个说话人每人20个汉语词的音视频语料.数据库的扩充与完善工作仍在不断的进行之中.在嘴唇检测方面,
我们对自适应色度过滤模型进行改进,提高了算法的鲁棒性,完成了对嘴唇的精确定位.结合DCT变换与K-L变换的各自特点,
我们提出了特征提取算法,使用较少维数的特征完成了对嘴唇区域主要信息的刻画.由于唇动信息同时包含了生理特征与行为特征,
我们使用静念动念混合建模的方式,完成了对说话人唇动个性特点的精确描述.在HMM训练时,我们提出了特征的归一化处理方法,提高了HMM在实际应用中的性能.
最后,我们分别对身份辨认系统与身份确认系统的基本理论进行了叙述,并完成了系统的实践工作.
关 键 词：身份识别 唇动 特征提取 隐马尔可夫模型 K-L变换 -paper completed the lip movements identification technology several basic issues of theoretical study, system as a whole be achieved. As this paper, the experimental basis, We have established a dynamic manner lip identification database (HITLUDB) The library currently contains 30 words each of 20 Chinese words, sound and video corpus. and the expansion of the database is still perfect keep going on. Detection of the lips, we adaptive color filter model improvements, improve the robustness of the algorithm, completed a pair of lips the precise positioning. DCT combined with the K-L transform their own characteristics, We have proposed a feature extraction algorithm, use less dimension of the lips completed the main message of regional characterization. As the lip movements of information, Platform: |
Size: 5222806 |
Author:李岚熙 |
Hits:
Description: 本文完成了对唇动身份识别技术几个基本问题的理论研究,并对整个系统加以实现.作为本文研究的实验基础,我们建立了唇动方式身份识别数据库(HITLUDB),
该库目前包含30个说话人每人20个汉语词的音视频语料.数据库的扩充与完善工作仍在不断的进行之中.在嘴唇检测方面,
我们对自适应色度过滤模型进行改进,提高了算法的鲁棒性,完成了对嘴唇的精确定位.结合DCT变换与K-L变换的各自特点,
我们提出了特征提取算法,使用较少维数的特征完成了对嘴唇区域主要信息的刻画.由于唇动信息同时包含了生理特征与行为特征,
我们使用静念动念混合建模的方式,完成了对说话人唇动个性特点的精确描述.在HMM训练时,我们提出了特征的归一化处理方法,提高了HMM在实际应用中的性能.
最后,我们分别对身份辨认系统与身份确认系统的基本理论进行了叙述,并完成了系统的实践工作.
关 键 词：身份识别 唇动 特征提取 隐马尔可夫模型 K-L变换 -paper completed the lip movements identification technology several basic issues of theoretical study, system as a whole be achieved. As this paper, the experimental basis, We have established a dynamic manner lip identification database (HITLUDB) The library currently contains 30 words each of 20 Chinese words, sound and video corpus. and the expansion of the database is still perfect keep going on. Detection of the lips, we adaptive color filter model improvements, improve the robustness of the algorithm, completed a pair of lips the precise positioning. DCT combined with the K-L transform their own characteristics, We have proposed a feature extraction algorithm, use less dimension of the lips completed the main message of regional characterization. As the lip movements of information, Platform: |
Size: 5222400 |
Author:QHLee |
Hits:
Description: DCT用于掌纹识别的word论文文档，比pdf更容易阅读编辑和引用，供大家学习和交流-The discrete cosine transform is a classical image processing technique, and the discriminant analysis is a commonly used image feature extraction technique. These two techniques are effectively combined in the paper, and a new palmprint feature extraction approach is proposed. After performing discrete cosine transform on the palmprint images, the proposed approach first uses a two-dimensional separability judgment to select the frequency bands with favorable separability. Then from the selected bands, it extracts the discriminant features by using an improved Fisherface method. The experimental results on the public palmprint image database demonstrate the effectiveness of the proposed approach. Platform: |
Size: 49152 |
Author:小哈 |
Hits:
Description: DCT域中MPEG7主色描述符的提取,该文在MPEG7的基础上提出了DCT域内直接提取主色描述符的新方法.这种方法节省了
对图像的解压缩的过程，因而大大的提高了对于压缩图像进行特征提取的速度和效果.作为整个算法的一部
分，一种自动阂值提取的算法也在该文中给予了描述.这种方法可以减少因人为设定经验闭值而带来的不确
定性，使算法更具鲁棒性.对比检索试验结果也说明本算法是一个高速有效的算法.新算法主要用于压缩图
像库或互联网上的相似检索.
-DCT domain MPEG7 Dominant Color descriptor extraction, the article at the basis of MPEG7 proposed DCT domain directly from the main color of the new method descriptor. This method saves the image in the process of decompression, which greatly enhance the compressed image for feature extraction speed and effectiveness. as part of the whole algorithm, an automatic threshold algorithm for extracting the text is also given to the description. This method can reduce the threshold to set the experience brought about by non- uncertainty, so that a more robust algorithm. Comparison Search results also shows that the algorithm is a fast and effective algorithm. The new algorithm used to compress the image database or a similar search on the Internet. Platform: |
Size: 403456 |
Author:杨康 |
Hits:
Description: matlab图像处理工具相，使用了主成分分析，ANN，SVM等方法。-This toolBox used in the image processing(feature extraction and classification)
PCA,LDA,ICA,DCT,RBF,RBE,GRNN,KNN,minimum distance,SVM, and others Platform: |
Size: 74752 |
Author:大长今 |
Hits:
Description: AR人脸库进行DCT变换，然后使用Fisher discriminant analysis 进行特征提取，使用cos分类器进行人脸分类。-AR Face Database for DCT transform, and then use the Fisher discriminant analysis feature extraction, using cos classifier for human face classification. Platform: |
Size: 2048 |
Author:邵珠立 |
Hits:
Description: 本文设计基于DCT的人脸识别系统，首先结合当今人脸识别的背景和发展状况讨论了人脸识别的研究内容及在各方面的应用；然后研究了人脸识别进行预处理，讨论了人脸识别预处理的其他方法，分析各种方法的利弊，最后采用DCT（离散余弦变换）实现人脸图像预处理中的降维处理；接下来对人脸图像的特征提取进行了研究，简单叙述了几何特征提取和代数特征提取，同时深入研究了基于DCT和PCA变换的人脸图像特征提取，从而实现是否对人脸识别系统识别率有所提高的研究；对于分类器的选择，本文对两种分类器进行了探讨，即最近邻分类器和BP神经网络分类器，同时采用BP神经网络分类器作为本次基于DCT的人脸识别系统设计的分类器，并对BP神经网络进行分类的算法进行设计，BP神经网络具有学习功能，只要采用本系统对训练图片进行训练，就可以记下图像的相关信息，对于测试图片，就可以很准确的识别出该图片是属于哪个的。最后，本文对整个人脸识别系统设计实验进行了实验分析，实验结果表明本文采用的方法切实有效。- This is the design of the DCT-based face recognition systems. First of all, the background light of the current face recognition and face recognition to discuss the development of research in all aspects of content and applications And then studied the pretreatment of face recognition to discuss the pre-treatment of other face recognition methods,and analysis of the pros and cons of various methods, finally, the use of DCT (discrete cosine transform) image pre-processing to achieve in the face of the reduced-order processing Next on the face image feature extraction have been studied, A brief description of the feature extraction and algebraic geometry feature extraction, while in-depth study based on the DCT and the PCA Transform face image feature extraction in order to achieve face recognition system to identify whether the rate of increase in the research For the choice of classifier, this paper carried out on two of classifier, that is, nearest neighbor classifier and BP neura Platform: |
Size: 422912 |
Author:刘文珍 |
Hits:
Description: 在DCT域进行图像检索，分别采取了对DCT块的边缘类型提取以及纹理特征提取两个方法-Image Retrieval in DCT domain, DCT blocks were taken on the edge of the type and texture feature extraction method of extracting the two Platform: |
Size: 898048 |
Author:liulu |
Hits:
Description: 先用小波变换进行降维后，再用DCT进行特征提取，然后用SVM分类识别，SVM需先安用libsvm工具箱，然后再可以运行，该程序包含ROL人脸库，一并上传。-First reduce the dimension of the wavelet transform, the then DCT feature extraction, and then use SVM classification, SVM must be safe to use libsvm toolbox, and then you can run, the program contains ROL face database, be uploaded. Platform: |
Size: 5598208 |
Author: |
Hits:
Description: 本文介绍了 DSP6711的硬件特性 分析了人脸检测、 识别的原理及算法的选型 运用基于 DCT变换域的 LDA的特征提取方法 ,实现了人脸的自动识别。在 Yale人脸库上的实验结果表明本算法识别率要比直接用 PCA进行特征提取的方
法要好-This article describes the DSP6711 hardware features analysis of face detection, recognition of the principle and algorithm selection the use of the LDA transform domain DCT-based feature extraction method, to achieve automatic recognition of human faces. In the Yale face database of experimental results show that the algorithm is the recognition rate than the direct use of PCA for feature extraction method is better Platform: |
Size: 339968 |
Author:louc |
Hits:
Description: 结合DCT和NNC进行人脸识别。先利用DCT提取特征，然后利用最近邻分类器分类，在ORL人脸库上测试效果不错。-The combination of DCT and NNC for face recognition. First DCT Feature Extraction, and then use the nearest neighbor classifiers, good test results on the ORL face database. Platform: |
Size: 1024 |
Author:尹贺峰 |
Hits:
Description: 结合DCT和BP神经网络进行人脸识别。先利用DCT提取特征，然后利用BP神经网络分类，在ORL人脸库上测试效果不错。-The combination of DCT and BP neural network for face recognition. First DCT Feature Extraction, and then use the BP neural network classifier, a good test results on the ORL face database. Platform: |
Size: 2048 |
Author:尹贺峰 |
Hits:
Description: 结合DCT和概率神经网络进行人脸识别。先利用DCT提取特征，然后利用PNN分类，在ORL人脸库上测试效果不错。-The combination of DCT and probabilistic neural network for face recognition. First DCT Feature Extraction, and then use a PNN classification, good test results on the ORL face database. Platform: |
Size: 1024 |
Author:尹贺峰 |
Hits:
Description: DWT and DCT based Robust Iris Feature Extraction and Recognition Algorithm for Biometric Personal Identification Platform: |
Size: 404480 |
Author:mhmd |
Hits:
Description: 提出了一种基于DCT提取人脸特征技术和支持向量机分类模型的人脸识别方法。利用离
散余弦变换可提取人脸可识别的大部分信息,而支持向量机作为分类器,在处理小样本、高维数等
方面具有独特的优势,且泛化能力很强,无需先验知识。从ORL 人脸库上的实验结果可以看出,
DCT特征提取是很有效的,且SVM的分类性能优于最近邻分类器,同时提高了整个系统的运算速
度。-A face recognition method based on DCT for face feature extraction and support vector machine classification model. Can extract most of the information face recognition using discrete cosine transform and support vector machine as classifier, has unique advantages in dealing with small sample, high dimension and generalization ability, without prior knowledge. As can be seen from the experimental results on the ORL database DCT feature extraction is very effective, and the SVM classification performance better than the nearest neighbor classifier, while increasing the speed of operation of the entire system. Platform: |
Size: 354304 |
Author:罗朝辉 |
Hits:
Search - dct feature extraction