An Embedded Application for Degraded Text Recognition
- PDF / 1,887,773 Bytes
- 9 Pages / 600 x 792 pts Page_size
- 61 Downloads / 227 Views
An Embedded Application for Degraded Text Recognition ´ Celine Thillou Laboratoire de Th´eorie des Circuits et Traitement du Signal, Facult´e Polytechnique de Mons, Bˆatiment Multitel-Initialis, 1 avenue Copernic, 7000 Mons, Belgium Email: [email protected]
Silvio Ferreira Laboratoire de Th´eorie des Circuits et Traitement du Signal, Facult´e Polytechnique de Mons, Bˆatiment Multitel-Initialis, 1 avenue Copernic, 7000 Mons, Belgium Email: [email protected]
Bernard Gosselin Laboratoire de Th´eorie des Circuits et Traitement du Signal, Facult´e Polytechnique de Mons, Bˆatiment Multitel-Initialis, 1 avenue Copernic, 7000 Mons, Belgium Email: [email protected] Received 24 December 2003; Revised 30 November 2004 This paper describes a mobile device which tries to give the blind or visually impaired access to text information. Three key technologies are required for this system: text detection, optical character recognition, and speech synthesis. Blind users and the mobile environment imply two strong constraints. First, pictures will be taken without control on camera settings and a priori information on text (font or size) and background. The second issue is to link several techniques together with an optimal compromise between computational constraints and recognition efficiency. We will present the overall description of the system from text detection to OCR error correction. Keywords and phrases: text detection, thresholding, character recognition, error correction.
1. INTRODUCTION A broad range of new applications and opportunities are emerging as wireless communication, mobile devices and camera technologies are becoming widely available and acceptable. These mature technologies introduce new research areas. One of the most fascinating frontier projects in the field of artificial intelligence is machine understanding of text. Extensive efforts have been made in order to give the blind or visually impaired access to text information; two complementary approaches are generally used. The first approach directly adapts the information support to the degree of blindness, by using either an optical zooming device that expands the character or Braille language. These solutions are not perfect. On one hand, optical enhancement solutions are cumbersome and not applicable in all cases. On the other hand, Braille language requires a complex learning and by the fact most of blind people do not know it. The second method consists in transforming textual information into
speech information. Some solutions combining a scanner, a pair of loudspeakers, and a computer currently exist. In addition to this material, the computer must be equipped with OCR, optical character recognition, and TTS, text-to-speech, technologies. OCR software aims at converting images from the scanner into text information while TTS software converts text information into a speech signal. This method has proved to be efficient with paper documents but presents the inconveniences of being limited to home use and to be
Data Loading...
