Abstract not found

For block-based classification, an image is divided into blocks and a feature vector is formed for each block by grouping statistics extracted from the block. Conventional block-based classification algorithms decide the class of a block by examining only the feature vector of this block and ignoring context information. In order to improve classi cation by context, an algorithm is proposed, which models images by two dimensional hidden Markov models (HMMs). The HMM considers feature vectors statistically dependent through an underlying state process assumed to be a Markov mesh, which has transition probabilities conditioned on the states of neighboring blocks from both horizontal and vertical directions. Thus, the dependency in two dimensions is reflected simultaneously. The HMM parameters are estimated by the EM algorithm. To classify an image, the classes with maximum a posteriori probability are searched jointly for all the blocks. Applications of the HMM algorithm to document and aerial image segmentation show that the algorithm outperforms CART TM,LVQ, and Bayes VQ.

This paper treats a multiresolution hidden Markov model for classifying images. Each image is represented by feature vectors at several resolutions, which are statistically dependent as modeled by the underlying state process, a multiscale Markov mesh. Unknowns in the model are estimated by maximum likelihood, in particular by employing the expectation-maximization algorithm. An image is classified by finding the optimal set of states with maximum a posteriori probability. States are then mapped into classes. The multiresolution model enables multiscale information about context to be incorporated into classification. Suboptimal algorithms based on the model provide progressive classification that is much faster than the algorithm based on single-resolution hidden Markov models.

We present an algorithm for automatic image orientation estimation using a Bayesian learning framework. We demonstrate that a small codebook (the optimal size of codebook is selected using a modified MDL criterion) extracted from a vector quantizer can be used to estimate the class-conditional densities of the observed features needed for the Bayesian methodology. We further show how feature clustering can be used as a feature selection mechanism to remove redundancies in the highdimensional feature vectors used for classification. Experiments on a database of 17; 901 images have shown that our proposed algorithm achieves an accuracy of approximately 97% on the training set and over 89% on an independent test set. 1. Introduction Content-based image organization and retrieval has emerged as an important area in computer vision and multimedia computing, due to the technological advances in digital imaging, storage, and networking. With the development of digital photography as well as...

In this paper, an algorithm is developed for segmenting document images into four classes: background, photograph, text, and graph. Features used for classi#cation are based on the distribution patterns of wavelet coe#cients in high frequency bands. Two important attributes of the algorithm are its multiscale nature---it classi#es an image at di#erent resolutions adaptively, enabling accurate classi#cation at class boundaries as well as fast classi#cation overall--- and its use of accumulated context information for improving classi#cation accuracy. Keywords document image segmentation, text and photograph segmentation, multiscale classi#cation, context-dependent classi#cation, wavelet transform, goodness of match. I. Introduction The image classi#cation problem considered in this paper is the segmentation of an image into four classes: background, photograph, text, and graph. Photograph refers to continuous-tone images, such as scanned pictures. Text is interpreted in the na...

Clustering applications cover several elds such as audio and video data compression, pattern recognition, computer vision, medical image recognition, etc. In this paper we present a new clustering algorithm called Enhanced LBG (ELBG). It belongs to the hard and K-means vector quantization groups and derives directly from the simpler LBG. The basic idea we have developed is the concept of utility of a codeword, a powerful instrument to overcome one of the main drawbacks of clustering algorithms: generally, the results achieved are not good in the case of a bad choice of the initial codebook. We will present our experimental results showing that ELBG is able to nd better codebooks than previous clustering techniques and the computational complexity is virtually the same as the simpler LBG.

A vector quantization algorithm is presented which accomplishes image restoration concurrently with image compression. The algorithm is based on nonlinear interpolative vector quantization (NLIVQ). An efficient codebook design procedure is also presented. A theoretical discussion of the algorithm is included along with results from simulations. 1. Introduction Vector quantization (VQ) is another name for what Shannon called block source coding subject to a fidelity criterion [1] . Coding of this type maps consecutive, usually non-overlapping, segments of input data to their best matching entry in a codebook of reproduction vectors. In the context of image coding, VQ is generally considered a data compression technique. However, VQ algorithms have been presented which perform other signal processing tasks concurrently with compression. These span the range from speech processing tasks such as speaker recognition and noise suppression, to image processing tasks like half-toning, edge det...

The connection between compression and the estimation of probability distributions has long been known for the case of discrete alphabet sources and lossless coding. A universal lossless code which does a good job of compressing must implicitly also do a good job of modeling. In particular, with a collection of codebooks, one for each possible class or model, if codewords are chosen from among the ensemble of codebooks so as to minimize bit rate, then the codebook selected provides an implicit estimate of the underlying class. Less is known about the corresponding connections between lossy compression and continuous sources. Here we consider aspects of estimating conditional and unconditional densities in conjunction with Bayes-risk weighted vector quantization for joint compression and classification.

Abstract—Combined compression and classification problems are becoming increasingly important in many applications with large amounts of sensory data and large sets of classes. These applications range from automatic target recognition (ATR) to medical diagnosis, speech recognition, and fault detection and identification in manufacturing systems. In this paper, we develop and analyze a learning vector quantization (LVQ) based algorithm for combined compression and classification. We show convergence of the algorithm using the ODE method from stochastic approximation. We illustrate the performance of our algorithm with some examples. Index Terms — Classification, compression, learning vector quantization, nonparametric, stochastic approximation.

This paper addresses the issue of optimal design of transform and multirate coders, when the compressed data are to be used for classification purposes. We propose to use Chernoff bounds on the probability of misclassification as the criterion for designing the coder. The performance of optimized transform and multirate coders is compared. The theory is illustrated with a texture classification example.