Design And Implementation Of Blind Reader Based On FPGA

In current society, more than99%text information appears in the form of visual paper materials. But the blind and visually impaired people can not read books and newspapers like the normal. Because of visual disability, they can not get information intuitively. A blind reader is just a device, which converts visual printed materials into sound. It makes people with visual impairments get paper text information conveniently without outsiders’ help. The system based on FPGA (Field Programmable Gate Array) can meet the needs of high speed image processing with its hardware features. What’s more, SOPC (System On a Programmable Chip) technology can make the design more flexible, and achieve software and hardware in-system programming and update.This paper designs a blind reader based on FPGA. The system uses Altera DE2development board as the hardware platform. With a CMOS (Complementary Metal Oxide Semiconductor) image sensor and a VGA (Video Graphic Array) device, it can collect printed character materials and display images both collected and processed. When the corresponding toggle switches are snapped, the system will output voice information via a speaker.The whole system consists of four modules, which are image capture module, image cache module, image display module and character recognition module. It adopts a synergistic manner through software and hardware. The image capture module, image cache module and image display module are written in Verilog hardware description language. The character recognition module is designed by SOPC Builder and programmed in C language in Nios Ⅱ IDE. Among them, the character recognition module is the core of the system, which decides the performance of the system. Text image processing is divided into preprocessing, feature extraction and matching identification. The paper discussed and analyzed the selection of algorithms in each part. The system adopts a median filter to smooth the gray level image firstly, and then it uses Otsu algorithm to convert the image into a binary one. After segmentation and normalization, it applies an improved thinning algorithm based on8neighborhoods to extract the character skeleton. Then the system extracts characters’feature points including endpoints, ambiguous points, turning points and four fork points, and it uses absolute distance measurement to match the characters’ feature vectors with the standard ones in the dictionary. Finally, it finds the corresponding audio files in the voice library, and the voice information will be output from a speaker. After repeated debugging and improvement, the system can achieve the functions of a blind reader.