CAJ | 학술논문

介绍基于ARM芯片JZ4755实现人眼疲劳状态监测的系统，该系统的算法过程是：通过8051单片机控制摄像头上不同每外波长LED灯的切换交替采集具有亮瞳效应的驾驶员眼睛状态的图像，经相临帧间的图像的中值滤波以及差分后根据瞳孔区域的明显特征来粗定位人眼位置，定位的方法用灰度积分投影。提取人眼小区域后再用最大类间方差法对原图像进行二值化处理，再用模版匹配的方法提取人眼的边界，最后根据PERCLOS及相关参数判驾驶员的疲劳状况。通过matlab的实验并在硬件上实现后的结果表明，该系统实用，算法简单快捷，而且由于是基于红外设计能全天监测驾驶员的疲劳状况。
개소기우ARM심편JZ4755실현인안피로상태감측적계통，해계통적산법과정시：통과8051단편궤공제섭상두상불동매외파장LED등적절환교체채집구유량동효응적가사원안정상태적도상，경상림정간적도상적중치려파이급차분후근거동공구역적명현특정래조정위인안위치，정위적방법용회도적분투영。제취인안소구역후재용최대류간방차법대원도상진행이치화처리，재용모판필배적방법제취인안적변계，최후근거PERCLOS급상관삼수판가사원적피로상황。통과matlab적실험병재경건상실현후적결과표명，해계통실용，산법간단쾌첩，이차유우시기우홍외설계능전천감측가사원적피로상황。
This paper is introduced eye fatigue monitoring system which is base on ARM chip of JZ4755, The process of the algorithm is ：The images with bright pupil effect of the drives＇ eyes is alternately collected by the different wavelengths＇ led lights on the camera, which is controlled by the MCU 8051.After the adjacent two images is processed by median filtering and differential characteristics, it can be rough positioning of the drivers＇ eyes by the obviously of the area of the pupil. And the method of rough positioning is gray integral projection. After getting the small area of the eyes using Otsu method on the original binary image processing, and then template matching method to extract the boundaries of the eyes, Finally ,it get the condition of fatigue in drivers by PERCLOS and related parameters. Through MATLAB experiments and hardware to achieve the results show that the system＇ s algorithm is not only practical, simple and rapid, but also it can be infrared the drivers＇ fatigue throughout the day because it is designed by infrared technology.