الفهرس 
AbstractOnly 14 pages are availabe for public view
 |  | from | 98 |  |  |
| | | from | 98 | | |
Abstract
Text detection and recognition in natural scene images received more interest in last years. It still an unsolved problem because some difficulties, such as some images contain complex background and low contrast, noise, various orientation styles and text can be of different font types and sizes. These difficulties make the automatic text extraction and recognizing it very difficult. This thesis proposes the implementation of an intelligent system for automatic detection of text from images and extracts and recognizes text in natural scene images by using some text detection algorithms to enhance text recognition. The proposed system implements various algorithms such as: Maximally Stable Extremal Regions (MSER) algorithm to detect the regions in the image. Canny edges algorithm to enhance edge detection. Connected Component Analysis algorithm (CCA) is used to determine the text components in the image. Stroke Width Transform algorithm (SWT) is used to determine the text pixels and exclude non-text regions. Bounding Box algorithm to detect and segment area of interest. Once extracting text from the image, the recognition process is done using Optical character Recognition(OCR). The system is evaluated using public datasets (ICDAR2003). The experimental results proved the robust performance of the proposed system.