Results 21  30
of
52
An Adaptive Algorithm for the Automatic Segmentation of Printed Arabic Text
"... ABSTRACT. Character segmentation is a crucial step in most Arabic optical text recognition systems. The recognition process depends mainly on the accuracy of the character segmentation. This paper presents a novel adaptive algorithm for the offline segmentation of printed Arabic text. There are man ..."
Abstract

Cited by 1 (0 self)
 Add to MetaCart
(Show Context)
ABSTRACT. Character segmentation is a crucial step in most Arabic optical text recognition systems. The recognition process depends mainly on the accuracy of the character segmentation. This paper presents a novel adaptive algorithm for the offline segmentation of printed Arabic text. There are many challenging features in the Arabic writing, for example, it is cursive and characters in a word can take four different shapes. A general rule cannot apply for segmenting all the characters. Hence, we propose an adaptive rulebased segmentation algorithm based on the general structural relationship of the Arabic text. The main rule used is that “most characters start with and end before a Tjunction on the baseline.” Some other rules are included to take care of the variations in the main rule. Results show the efficiency of the proposed algorithm, where it is found to achieve a segmentation accuracy of 96.5%. 1.
Stroke Segmentation from Livestock Brand Images
, 2009
"... Abstract. The detection and extraction of primitive curves are important stages in line analysis systems used to handle considerable number of digital image processing problems. This article presents a stroke segmentation approach suitable to perform similarity measuring, using digital image process ..."
Abstract

Cited by 1 (1 self)
 Add to MetaCart
Abstract. The detection and extraction of primitive curves are important stages in line analysis systems used to handle considerable number of digital image processing problems. This article presents a stroke segmentation approach suitable to perform similarity measuring, using digital image processing techniques, in a cattle brand registration system. The skeletons of the brands are analyzed to detect and separate primary strokes at junction and intersection points. Primary strokes are then mapped into a threedimensional orientation space to group them in the primitive continuous strokes. The results of a set of experiences are presented.
UNSUPERVISED LEARNING: AN INFORMATION THEORETIC FRAMEWORK By
, 2008
"... I would like to take this opportunity to thank my advisor Dr. José C. Príncipe for his constant, unwavering support and guidance throughout my stay at CNEL. He has been a great mentor pulling me out of many local minima (in the language of CNEL!). I still wonder how he works nonstop from morning to ..."
Abstract

Cited by 1 (0 self)
 Add to MetaCart
I would like to take this opportunity to thank my advisor Dr. José C. Príncipe for his constant, unwavering support and guidance throughout my stay at CNEL. He has been a great mentor pulling me out of many local minima (in the language of CNEL!). I still wonder how he works nonstop from morning to evening without lunch, and I am sure this feeling is shared among many of my colleagues. In short, he has been an inspiration and continues to be so. I would like to express my gratitude to all my committee members; Dr. Murali Rao, Dr. John G. Harris and Dr.Clint Slatton; for readily agreeing to be part of my committee. They have helped immensely in improving this dissertation with their inquisitive nature and helpful feedbacks. I would like to especially thank Dr. Murali Rao, my math mentor, for keeping a vigil on all my notations and bringing sophistication to my engineering mind! Special mention is also needed for Julie, the research coordinator at CNEL, for constantly monitoring the pressure level at the lab and making us smile even if it is for a short while. My past as well as present colleagues at CNEL need due acknowledgement. Without them, I would have been shooting questions to walls and mirrors! I have grown to
Principal Curve Time Warping
"... AbstractTime warping finds use in many fields of time series analysis, and it has been effectively implemented in many different application areas. Rather than focusing on a particular application area we approach the general problem definition, and employ principal curves, a powerful machine lear ..."
Abstract
 Add to MetaCart
AbstractTime warping finds use in many fields of time series analysis, and it has been effectively implemented in many different application areas. Rather than focusing on a particular application area we approach the general problem definition, and employ principal curves, a powerful machine learning tool, to improve the noise robustness of existing time warping methods. The increasing noise level is the most important problem that leads to unnatural alignments. Therefore, we tested our approach in low signaltonoise ratio (SNR) signals, and obtained satisfactory results. Moreover, for the signals denoised by principal curve projections we propose a differential equationbased time warping method, which has a comparable performance with lower computational complexity than the existing techniques.
FHS Z;]Jƒ… „ ƒ q&e ƒW † nMETHOD AND SOFTWARE FOR FAST CONSTRUCTION OF PRINCIPAL MANIFOLDS APPROXIMATIONS
"... f#gJhNi P9KNRjQJRBk;lcU h K h R?I ..."
(Show Context)
Topological grammars for data approximation
, 2006
"... A method of topological grammars is proposed for multidimensional data approximation. For data with complex topology we define a principal cubic complex of low dimension and given complexity that gives the best approximation for the dataset. This complex is a generalization of linear and nonlinear ..."
Abstract
 Add to MetaCart
(Show Context)
A method of topological grammars is proposed for multidimensional data approximation. For data with complex topology we define a principal cubic complex of low dimension and given complexity that gives the best approximation for the dataset. This complex is a generalization of linear and nonlinear principal manifolds and includes them as particular cases. The problem of optimal principal complex construction is transformed into a series of minimization problems for quadratic functionals. These quadratic functionals have a physically transparent interpretation in terms of elastic energy. For the energy computation, the whole complex is represented as a system of nodes and springs. Topologically, the principal complex is a product of onedimensional continuums (represented by graphs), and the grammars describe how these continuums transform during the process of optimal complex construction. This factorization of the whole process onto onedimensional transformations using minimization of quadratic energy functionals allows us to construct efficient algorithms. c ○ 2006 Elsevier Ltd. All rights reserved.
Automatic Trajectory Extraction And Validation Of Scanned Handwritten Characters
"... A wellestablished task in forensic writer identification is the comparison of prototypical character shapes (allographs) present in the handwriting. Using elastic matching techniques like Dynamic Time Warping (DTW), comparison results can be made that are plausible and understandable to the human e ..."
Abstract
 Add to MetaCart
(Show Context)
A wellestablished task in forensic writer identification is the comparison of prototypical character shapes (allographs) present in the handwriting. Using elastic matching techniques like Dynamic Time Warping (DTW), comparison results can be made that are plausible and understandable to the human expert. Since these techniques require the dynamics of the handwritten trace, the “online” dynamic allograph trajectories need to be extracted from the “offline ” scanned documents. We have implemented an algorithm that can automatically extract this information from scanned images. The algorithm makes a list of all possible trajectories. Using a number of traditional techniques and DTW for evaluation, the best trajectory is selected. To be able to make a quantitative assessment of our techniques, rather than a qualitative discussion of a small number of examples, we tested the performance on two large datasets, one containing online and the other containing offline data. Two different methods (one for the online, and one for the offline dataset) are used to validate the generated trajectories. The results of the experiments show that DTW can significantly improve the performance of trajectory extraction when compared to traditional techniques.
Skeletonization of TwoDimensional regions Using Hybrid Method
"... Conversion of twodimensional objects into a skeletal representation forms an essential step in many image processing and pattern recognition applications. Most of the topological structure of objects, and the information contained in the outline of their shapes, are preserved in the skeleton. Appro ..."
Abstract
 Add to MetaCart
(Show Context)
Conversion of twodimensional objects into a skeletal representation forms an essential step in many image processing and pattern recognition applications. Most of the topological structure of objects, and the information contained in the outline of their shapes, are preserved in the skeleton. Approaches based on Voronoi techniques preserve topology, but heuristic measures are introduced to remove unwanted edges. Methods based on Euclidean distance functions can localize skeletal points accurately, but often at the cost of altering the topology of the object. In this paper we offer a method to generate skeletal representations combining these two methods, which is robust and accurate, and preserves topology.
COMPLEX CURVE TRACING BASED ON A MINIMUM SPANNING TREE MODEL AND REGULARIZED FUZZY CLUSTERING
"... The fuzzy curvetracing (FCT) algorithm can be used to extract a smooth curve from unordered noisy data. However, the model produces good results only if the curve shape is either opened or closed. In this paper, we propose several techniques to generalize the FCT algorithm for tracing complicated c ..."
Abstract
 Add to MetaCart
(Show Context)
The fuzzy curvetracing (FCT) algorithm can be used to extract a smooth curve from unordered noisy data. However, the model produces good results only if the curve shape is either opened or closed. In this paper, we propose several techniques to generalize the FCT algorithm for tracing complicated curves. We develop a modified clustering algorithm that can produce cluster centers less dependent on the prespecified number of clusters, which makes the reordering of cluster centers easier. We make use of the Eikonal equation and the Prim’s algorithm to form the initial curve, which may contain sharp corners and intersections. We also introduce a more powerful curve smoothing method. Our generalized FCT algorithm is able to trace a wide range of complicated curves, such as handwritten Chinese characters. 1.