Related Experiment Video
Updated: Jul 16, 2026
05:38
Interaction between Phonological and Semantic Processes in Visual Word Recognition using Electrophysiology
Published on: June 29, 2021
Recognition of pixelized Chinese characters using simulated prosthetic vision
1Department of Biomedical Engineering, Shanghai Jiaotong University, Shanghai, China.
Artificial Organs
|March 9, 2007
Summary
Restoring reading for the blind requires optimizing Chinese character recognition. A minimum of 12x12 pixels and 150 electrodes are recommended for visual prostheses, with simpler fonts and fewer strokes improving accuracy.
Area of Science:
- Biomedical Engineering
- Neuroscience
- Computer Vision
Background:
- Visual prostheses aim to restore reading ability for the blind.
- Limited stimulating electrodes pose a challenge for high-resolution visual information.
- Chinese character recognition is complex due to intricate structures.
Purpose of the Study:
- To determine the optimal pixel resolution for Chinese character recognition by the blind.
- To inform the design of visual prostheses for effective reading rehabilitation.
- To identify key factors influencing character recognition accuracy.
Main Methods:
- Developed the HanziConvertor platform for digital image processing.
- Tested 40 normal-sighted subjects using pixelized Chinese character patterns.
- Analyzed recognition accuracy based on pixel number, font type, and stroke count.
Main Results:
- Optimal recognition accuracy achieved with at least 12x12 binary pixels.
- Recommended a minimum of 150 electrodes for visual prostheses.
- Song Ti and Hei Ti fonts demonstrated superior clarity and effectiveness.
- Characters with fewer strokes yielded higher recognition accuracy.
Conclusions:
- A 12x12 pixel array is crucial for effective Chinese character recognition in visual prostheses.
- Font choice and character complexity significantly impact reading rehabilitation.
- The findings provide essential data for developing advanced visual prostheses for the blind.
