基于深度学习的视力保护系统

Vision Protection System Based on Deep Learning

针对目前普遍存在的用眼过度、不良坐姿、光线过强(过弱)等问题导致的视力下降现象,设计研发了一套基于深度学习技术的多功能视力保护系统。在上位机上依据Tensorflow框架训练自己的数据集实现了YOLOv3的算法程序来进行用户阅读状态的检测。在下位机上外接光敏电阻检测光照强度,红外传感器检测人体距离,当用户距离书桌小于15 cm或大于50 cm时驱动电机控制椅背的角度。使用Android Studio开发良好交互界面的APP,采用MYSQL建立用户使用信息、用眼记录的数据库。所设计的视力保护系统,能为用户提供适宜的用眼环境,有效的保护了视力。

A multi-functional vision protection system based on deep learning technology was designed and developed to solve the common visual impairment caused by excessive eye use,poor posture,and too strong(too weak)light.The algorithm program of YOLOv3 was implemented on the upper computer according to the Tensorflow framework to train the data set to detect the reading state of users.The external photosensitive resistance of the next machine detects the light intensity,and the infrared sensor detects the distance between the human body.When the distance between the user and the desk is less than 25cm or more than 50cm,the driving motor controls the Angle of the back of the chair.Android Studio was used to develop apps with good interactive interfaces,and MYSQL was used to establish a database of user usage information and eye-recording.The designed visual acuity protection system provides the user with a suitable eye environment and effectively protects the visual acuity.