A Study on a Goal Oriented Detection and Verification Based Approach for Image and Ink Document Analysis Zhenlong Bai audiobook

A Study on a Goal Oriented Detection and Verification Based Approach for Image and Ink Document Analysis Zhenlong Bai audiobook

A Study on a Goal Oriented Detection and Verification Based Approach for Image and Ink Document Analysis Zhenlong Bai audiobook free

This dissertation, "A Study on a Goal Oriented Detection and Verification Based Approach for Image and Ink Document Analysis" by Zhenlong, Bai, 白真龍, was obtained from The University of Hong Kong (Pokfulam, Hong Kong) and is being sold pursuant to Creative Commons: Attribution 3. 0 Hong Kong License. The content of this dissertation has not been altered in any way. We have altered the formatting in order to facilitate the ease of printing and reading of the dissertation. All rights not granted by the above license are retained by the author. Abstract: Abstract of thesis entitled "A Study on A Goal Oriented Detection and Verification Based Approach for Image and Ink Document Analysis" Submitted by Zhenlong BAI for the degree of Doctor of Philosophy at The University of Hong Kong in December 2005 Documentimageandinkanalysisistocorrectlygetgeometricpartitionandobtain useful text information. Although there are many existing algorithms, there are still two challenges when applying them into practical applications, namely: 1). To be with greater generalization, which is favored by various new applications, 2). To be more reliable, which is favored by task-dependent and human-computer interactionsystems. Inthisthesis, westudyagoal-orienteddetectionandverifica- tion based approach to handle these two challenges, and implement this approach in different applications. The process of analyzing a document can be regarded as a procedure of de- tecting different geometric objects in the document. Those geometric objects include text lines, text blocks, underlines, words, etc. In our approach, the goal is first identified, then all possible geometric objects are figured out according to the goal and added into the detection space. Then, a series of processing stages are invoked adaptively in a sequential scheme, each includes a detection module and a verification module for a certain kind of geometric objects. At each stage, detection results that pass the following verification module are output, other- wise, those with low confidence will go to next stage for further processing. In this way, each stage is guided by the result of a preceding stage, and the detection space is reduced step by step, thus later detections are more efficient and more reliable. This scheme is also a good way for multi-strategies combination. The whole procedure ends when the goal is reached or all known strategies have been tried. In this sequential scheme, geometric objects with higher confidences that can be verified are processed in earlier stages. In our studies, four practical applications are applied to verify the efficacy of this approach. First, a single-stage scheme of this approach is adopted for Chi- nese business card image analysis, which shows the advantage of "goal-oriented analysis" when the goal can be well defined. Then a three-stage scheme is applied for pen scanner image analysis, which demonstrates the flexibility of adopting our approach for a new application and for a human-machine interaction system. After that, a five-stage scheme is used for underline detection and removal. It verifiestheapproachforhighlytask-dependentandlessdocument-typeconcerned systems. Finally, this approach is applied to deal with online Chinese ink doc- uments, where the nature of the data is totally different. Because our approach depends greatly on character recognition results for verification, a new approach is proposed and developed for online handwritten Chinese character recognition by using 8-directional features. It is demonstrated that through above applications, our approach offers a good solution to addressing the above mentioned two challenges in image and ink document analysis. (445 words) DOI: 10. 5353/th_b3660000 Subjects: Image analysis - Data processingAlgorithms.

Author:

download book

A Study on a Goal Oriented Detection and Verification Based Approach for Image and Ink Document Analysis Zhenlong Bai audio book download

  • Author:
  • Publisher: Open Dissertation Press
  • Publication date:
  • Cover: Paperback
  • Language: English
  • ISBN-10: 1361434236
  • ISBN-13: 978-1361434239
  • Dimensions: 8.5 x 0.4 x 11 inches
  • Weight: 14.9 ounces
  • Hardcover:
  • Series:
  • Grade:
  • Age:
  • Author:
  • Price: $49.00

Book reviews

A Study on a Goal Oriented Detection and Verification Based Approach for Image and Ink Document Analysis

Book Title

Size

Link

A Study on a Goal Oriented Detection and Verification Based Approach for Image and Ink Document Analysis read from EasyFiles

4.5 mb. download book

A Study on a Goal Oriented Detection and Verification Based Approach for Image and Ink Document Analysis download from OpenShare

3.8 mb. download free

A Study on a Goal Oriented Detection and Verification Based Approach for Image and Ink Document Analysis download from WeUpload

4.8 mb. read book

A Study on a Goal Oriented Detection and Verification Based Approach for Image and Ink Document Analysis download from LiquidFile

5.9 mb. download

A Study on a Goal Oriented Detection and Verification Based Approach for Image and Ink Document Analysis Zhenlong Bai audio book download

Book Title

Size

Link

A Study on a Goal Oriented Detection and Verification Based Approach for Image and Ink Document Analysis read in djvu

5.2 mb. download DjVu

A Study on a Goal Oriented Detection and Verification Based Approach for Image and Ink Document Analysis download in pdf

5.7 mb. download Pdf

A Study on a Goal Oriented Detection and Verification Based Approach for Image and Ink Document Analysis download in odf

3.2 mb. download ODF

A Study on a Goal Oriented Detection and Verification Based Approach for Image and Ink Document Analysis download in epub

4.5 mb. download ePub

Related books