混合模式识别系统研究

讨论基于多种分类方法的模块组合实现的混合模式识别系统，它不同于利用多分类器输出结果表决的集成系统。提出两个系统：一个面向刷体汉字文本识别，另一个面策自由手写体字识别。     

Recognition of Handwritten ZIP Codes in a Real—World Non-Standard-Letter Sorting System

In this article, we describe the OCR and image processing algorithms used to read destination addresses from non-standard letters (flats) by Siemens postal automation system currently in use by the Deutsche Post AG¹.We first describe the sorting machine, its OCR hardware and the sequence of image processing and pattern recognition algorithms needed to solve the difficult task of reading mail addresses, especially handwritten ones. The article concentrates mainly on the two classifiers used to recognize handprinted digits. One of them is a complex time delayed neural network (TDNN) used to classify scaled digit-features. The other classifier extracts the structure of each digit and matches it to a number of prototypes. Different digits represented by the same graph are then discriminated by classifiying some of the features of the digit-graph with small neural networks.We also describe some approaches for the segmentation of the digits in the ZIP code, so that the resulting parts can be processed and evaluated by the classifiers.     

A database for handwritten text recognition research

An image database for handwritten text recognition research is described. Digital images of approximately 5000 city names, 5000 state names, 10000 ZIP Codes, and 50000 alphanumeric characters are included. Each image was scanned from mail in a working post office at 300 pixels/in in 8-bit gray scale on a high-quality flat bed digitizer. The data were unconstrained for the writer, style, and method of preparation. These characteristics help overcome the limitations of earlier databases that contained only isolated characters or were prepared in a laboratory setting under prescribed circumstances. Also, the database is divided into explicit training and testing sets to facilitate the sharing of results among researchers as well as performance comparisons     

Robust vision-based features and classification schemes for off-line handwritten digit recognition

We use well-established results in biological vision to construct a model for handwritten digit recognition. We show empirically that the features extracted by our model are linearly separable over a large training set (MNIST). Using only a linear discriminant system on these features, our model is relatively simple yet outperforms other models on the same data set. In particular, the best result is obtained by applying triowise linear support vector machines with soft voting on vision-based features extracted from deslanted images.     

基于模版对比的手写体数字识别神经网络模型

针对现有手写体数字识别神经网络模型的不足,提出基于模版对比的改进方法。建立8×12像素的手写体数字0～9的标准模版,则模版中每个数字与其他数字之间存在一定的像素差异,以此作为标准模版差异值。由于书写存在不确定性,采用在一定范围内随机增大或减小标准模版差异值的方法来构建神经网络模型的训练样本、检验样本与测试样本。在遵循建模基本原则和步骤的情况下,建立了泛化能力较好的手写体数字识别的神经网络模型。实验表明：该方法建模便捷、实用性好,测试样本的正确识别率达99.6%以上。     

Reading handwritten digits: a ZIP code recognition system

A neural network algorithm-based system that reads handwritten ZIP codes appearing on real US mail is described. The system uses a recognition-based segmenter, that is a hybrid of connected-components analysis (CCA), vertical cuts, and a neural network recognizer. Connected components that are single digits are handled by CCA. CCs that are combined or dissected digits are handled by the vertical-cut segmenter. The four main stages of processing are preprocessing, in which noise is removed and the digits are deslanted, CCA segmentation and recognition, vertical-cut-point estimation and segmentation, and directly lookup. The system was trained and tested on approximately 10000 images, five- and nine-digit ZIP code fields taken from real mail     

A fuzzy logic system for the detection and recognition ofhandwritten street numbers

Fuzzy logic is applied to the problem of locating and reading street numbers in digital images of handwritten mail. A fuzzy rule-based system is defined that uses uncertain information provided by image processing and neural network-based character recognition modules to generate multiple hypotheses with associated confidence values for the location of the street number in an image of a handwritten address. The results of a blind test of the resultant system are presented to demonstrate the value of this new approach. The results are compared to those obtained using a neural network trained with backpropagation. The fuzzy logic system achieved higher performance rates     

Data driven image models through continuous joint alignment

This paper presents a family of techniques that we call congealing for modeling image classes from data. The idea is to start with a set of images and make them appear as similar as possible by removing variability along the known axes of variation. This technique can be used to eliminate "nuisance" variables such as affine deformations from handwritten digits or unwanted bias fields from magnetic resonance images. In addition to separating and modeling the latent images - i.e., the images without the nuisance variables - we can model the nuisance variables themselves, leading to factorized generative image models. When nuisance variable distributions are shared between classes, one can share the knowledge learned in one task with another task, leading to efficient learning. We demonstrate this process by building a handwritten digit classifier from just a single example of each class. In addition to applications in handwritten character recognition, we describe in detail the application of bias removal from magnetic resonance images. Unlike previous methods, we use a separate, nonparametric model for the intensity values at each pixel. This allows us to leverage the data from the MR images of different patients to remove bias from each other. Only very weak assumptions are made about the distributions of intensity values in the images. In addition to the digit and MR applications, we discuss a number of other uses of congealing and describe experiments about the robustness and consistency of the method.     

基于多分类器组合的手写体数字识别

本文提出了一个基于多分类器组合的手写体数字识别方法。文中首先给出了一个客观评价分类器性能的参数,其后基于此参数提出了多分类器的组合方法,并从理论上研究了此方法的一些性质,本文实验采用Ｃｏｎｃｏｒｔｄｉａ大学模式识别与机器智能中心的手写体数字数据库,在实验中,使用了９个利用不同特征分类器进行组合,组合后识别率、拒识率和可靠性分别可达到９７．０５％,２．０５％,９９．０８％。     

Recognizing handwritten digits using hierarchical products ofexperts

The product of experts learning procedure can discover a set of stochastic binary features that constitute a nonlinear generative model of handwritten images of digits. The quality of generative models learned in this way can be assessed by learning a separate model for each class of digit and then comparing the unnormalized probabilities of test images under the 10 different class-specific models. To improve discriminative performance, a hierarchy of separate models can be learned, for each digit class. Each model in the hierarchy learns a layer of binary feature detectors that model the probability distribution of vectors of activity of feature detectors in the layer below. The models in the hierarchy are trained sequentially and each model uses a layer of binary feature detectors to learn a generative model of the patterns of feature activities in the preceding layer. After training, each layer of feature detectors produces a separate, unnormalized log probability score. With three layers of feature detectors for each of the 10 digit classes, a test image produces 30 scores which can be used as inputs to a supervised, logistic classification network that is trained on separate data     

基于PCA初始化卷积核的CNN手写数字识别算法

针对卷积神经网络对手写数字识别训练在卷积核随机初始化情况下收敛速度慢和识别率低的问题，提出一种主成分分析（PCA）初始化卷积核的卷积神经网络（CNN）手写数字识别算法。算法首先选取训练样本集并将其送入CNN，在相应层对Feature Map进行全覆盖取图像块处理，然后进行分层PCA学习，将学习到的特征向量做为对应卷积层的卷积核参数进行初始化，最后再用这些卷积核对原始图像进行卷积操作。实验结果表明，与随机初始化卷积核的CNN手写数字识别算法相比，改进的算法在应用MNIST数据库训练时不仅收敛，而且在产生相同均方误差的情况下迭代次数少，识别率高。     

手写中文信封的地址行字符切分算法

在手写体中文信封处理系统中,地址行字符切分是实现地址行识别的关键步骤。本文根据邮政信封地址行字符的特点,有针对性的提出了一种字符切分算法。首先对地址行图像利用投影、求连通区域、笔划穿越数分析等基于字符结构的方法进行初始切分,得到基本字段序列;然后通过对相邻的基本字段进行组合形成多条候选切分路径,再通过识别的可信度和邮政目标地址库的先验知识信息对路径进行评价分析,从而得到最优的切分路径。该算法经过邮政分拣机采集的实际信封图像测试,纯地址行识别正确率达到78.61%,地址行识别与邮政编码识别相结合的分拣正确率达到95.42%。     

Lexicon-driven segmentation and recognition of handwritten character strings for Japanese address reading

This paper describes a handwritten character string recognition system for Japanese mail address reading on a very large vocabulary. The address phrases are recognized as a whole because there is no extra space between words. The lexicon contains 111,349 address phrases, which are stored in a trie structure. In recognition, the text line image is matched with the lexicon entries (phrases) to obtain reliable segmentation and retrieve valid address phrases. The paper first introduces some effective techniques for text line image preprocessing and presegmentation. In presegmentation, the text line image is separated into primitive segments by connected component analysis and touching pattern splitting based on contour shape analysis. In lexicon matching, consecutive segments are dynamically combined into candidate character patterns. An accurate character classifier is embedded in lexicon matching to select characters matched with a candidate pattern from a dynamic category set. A beam search strategy is used to control the lexicon matching so as to achieve real-time recognition. In experiments on 3,589 live mail images, the proposed method achieved correct rate of 83.68 percent while the error rate is less than 1 percent.     

卷积神经网络SIP微系统实现

近年来,微电子技术进入到纳电子/集成微系统时代,SIP（System in Package）和SOC（System on Chip）是微系统实现的两种重要技术途径;基于神经网络的深度学习技术在图形图像、计算机视觉和目标识别等方面得以广泛应用。卷积神经网络的深度学习技术在嵌入式平台的小型化、微型化是一项重要研究领域。如何将神经网络轻量化和微系统相结合,达到性能、体积和功耗的最优化平衡是一难点。介绍了一款将SIP技术和基于FPGA的卷积神经网络相结合的微系统实现方案,它以Zynq SOC和FLASH、DDR3存储器为主要组成,利用SIP高密度系统封装技术进行集成,在其中的PL端（FPGA）采用HLS来设计CNN（Convolutional Neural Network,卷积神经网络）中的卷积层和池化层,生成IP核,分时复用构建微系统,设计实现了Micro_VGGNet轻量化模型。测试采用MNIST手写数字数据集作为训练和测试样本,该微系统能够实准确识别手写数字,准确率达到98.1%。体积仅为30 mm×30 mm×1.2 mm,在100 MHz工作频率下,?图像处理速度可达到20.65 FPS,功耗仅为2.1 W,实现了轻量化神经网络微系统的多目标平衡（性能、体积和功耗）。     

基于动态模板匹配的移动机器人目标识别

将视觉显著性与基于动态模板匹配的目标识别方法相结合,提出了一种适用于未知环境下的移动机器人目标识别方法.具体而言,首先设计了基于分布式控制的移动机器人视觉系统,提高了视频处理效率;之后利用基于背景先验的显著性检测方法对图像进行预处理,排除了相对次要的背景区域;最后对处理后的图像进行动态模板匹配,提高了目标识别的准确率.实验结果表明,该方法能够满足移动机器人在目标识别过程中对图像处理的实时性和准确性的要求,具有良好的有效性.     

一种改进的区域双目立体匹配方法

双目立体匹配是机器视觉中的热点、难点问题。分析了区域立体匹配方法的优缺点,提出了改进的区域立体匹配方法。首先,采集双目视觉图像对对图像对进行校正、去噪等处理,利用颜色特征进行图像分割,再用一种快速有效的块立体匹配算法对图像进行立体匹配。然后,在匹配过程中使用绝对误差累积(SAD)的小窗口来寻找左右两幅图像之间的匹配点。最后,通过滤波得到最终的视差图。实验表明:该方法能够有效地解决重复区域、低纹理区域、纹理相似区域、遮挡区域等带来的误匹配问题,能得到准确清晰的稠密视差图。     

Shape matching using LAT and its application to handwritten numeralrecognition

This paper describes an iterative technique for gradually deforming a mask binary image with successive local affine transformation (LAT) operations so as to yield the best match to an input binary image as one new and promising approach toward robust handwritten character recognition. The method uses local shapes in the sense that the LAT of each point at one location is optimized using locations of other points by means of least-squares data fitting using Gaussian window functions. It also uses a multiscale refinement technique that decreases the spread of window functions with each iteration. Especially in handwritten character recognition, structural information is indispensable for robust shape matching or discrimination. The method is enhanced to explicitly incorporate structures by weighting the above least-squares criterion with similarity measures of both topological and geometric features of the mask and input images. Moreover, deformation constraints are imposed on each iteration, not only to promote and stabilize matching convergence but also to suppress an excessive matching process. Shape matching experiments have been successfully carried out using skeletons of totally unconstrained handwritten numerals     

基于DSP的手写数字识别系统

介绍了基于TMS320VC5402的手写数字识别系统和该系统的基本原理,给出了它的硬件原理图和软件设计程序框图。最后描述了所用的识别算法和改进的训练算法。