Practical Application of Intelligent Vision Measurement System Based on Deep Learning
10.12455/j.issn.1671-7104.230652
- VernacularTitle:基于深度学习的智能视力测量系统的应用实践
- Author:
Ruilin HU
1
;
Dan SUN
;
Guilian SHI
;
Anpeng PAN
Author Information
1. 温州医科大学附属眼视光医院,温州市,325027;国家眼耳鼻喉疾病(眼部疾病)临床医学研究中心,温州市,325027
- Keywords:
deep learning;
lateral blinking detection;
vision fluctuation;
visual function;
blink characteristics
- From:
Chinese Journal of Medical Instrumentation
2024;48(4):380-384,391
- CountryChina
- Language:Chinese
-
Abstract:
To comprehensively assess the true visual function of clinical dry eye patients and the comprehensive impact of blinking characteristics on functional vision of the human eye,an intelligent vision measurement system has been designed and developed to detect and analyze blinks from the side.The system employs deep learning keypoint recognition technology to analyze eyelid features from a lateral perspective.It presents the data of identified key points for the upper and lower eyelids in a line chart format and annotates the trough of each blink.By setting benchmark values,the system automatically calculates the proportion of complete and incomplete blinks in the tested individuals.The results indicate that the system is stable in performance and accurate in measurement,successfully achieving the anticipated design objectives.It thereby provides reliable technical support for future clinical applications.
