文献类型：期刊文献

中文题名：一种基于HSV和LBP特征融合的眼疲劳诊断方法

英文题名：Eye Fatigue Diagnosis Method Based on Feature Fusion by HSV and LBP

作者：李东[1];彭亦功[1]

机构：[1]华东理工大学信息科学与工程学院,上海200237

年份：2016

卷号：37

期号：10

起止页码：77

中文期刊名：自动化仪表

外文期刊名：Process Automation Instrumentation

收录：CSTPCD

基金：上海市重点学科建设基金资助项目(编号:B504)

语种：中文

中文关键词：特征提取;HSV;LBP;鲁棒性;特征融合;眼疲劳诊断;状态识别

外文关键词：Feature extraction HSV LBP Robustness Feature fusion Eye fatigue diagnosis State recognition

摘要：针对色调饱和度亮度(HSV)颜色空间特征提取方法在低光照条件下人眼状态识别率低、局部二值模式(LBP)特征提取方法在复杂光照下鲁棒性不强的问题,根据人眼在不确定光照情况下的状态识别需求,提出了基于HSV颜色特征和LBP纹理特征融合的眼疲劳诊断方法。仿真试验表明,该方法可有效提高单独使用HSV和LBP特征提取方法仿真时的人眼状态识别率和眼疲劳诊断精度,具有很好的眼疲劳诊断效果,应用前景广泛。
Under low light conditions,the color space feature extraction method of HSV （ Hue Saturation Value） provides low recognition rate of eye state; while under complex lighting conditions, the feature extraction method of LBP （Local Binary Pattern） features low robustness, in accordance with the demands for recognizing eye state under uncertain illumination conditions, the eye fatigue diagnosis method based on HSV and LBP, that fusing color feature and texture feature is proposed. Simulation experiments show that the proposed method can effectively improve eye state recognition rate and eye fatigue diagnostic accuracy obtained by using HSV or LBP separately. Besides, it has a good effect on eyefatigue diagnosis and a wide prospect in application.