虫蛀麦粒的太赫兹成像检测方法研究  

The Detection of Vermiculate Grain of Wheat Using Terahertz Imaging

作　　者：琚新刚 廉飞宇[2] 张元[1,2] 葛宏义[2] 蒋玉英 Ju Xingang;Lian Feiyu;Zhang Yuan;Ge Hongyi;Jiang Yuying(Henan Food Crops Collaborative Innovation CenterI,Zhengzhou 45004;Henan Province Key Laboratory of Grain Photoelectric Detection and Contro;Henan University of Technology2,Zhengzhou 45000;Circuits and Systems Key Disciplines of Henan Education Institute,Henan Institute of Education3,Zhengzho)

机构地区：[1]河南粮食作物协同创新中心,郑州450046 [2]河南省粮食光电探测与控制重点实验室,河南工业大学,郑州450001 [3]河南教育学院电路与系统重点学科,河南教育学院,郑州450046

出　　处：《中国粮油学报》2018年第8期106-111,118,共7页Journal of the Chinese Cereals and Oils Association

基　　金：河南省基础与前沿计划项目(152300410079)

摘　　要：针对小麦蛀食性害虫过去检测缺乏有效的技术手段问题,本研究将太赫兹成像检测应用到麦粒内部结构检测。首先,在太赫兹光谱仪加装反射成像模块,设置回波时延30~33 ps,对麦粒内部结构进行太赫兹层析成像;然后,针对麦粒内部结构太赫兹剖面图,采用最大类间方差法分割图像中的蛀食区域,并引入遗传算法,提高了最大类间方差法求解最佳阈值的效率,实验结果表明,本文的方法能够有效地识别出麦粒内部的虫蛀区域。Aiming at the lack of effective technique for early detection of boring pests in stored grain, we appliedterahertz imaging in detection of grain of wheat. First, we added a catoptric imaging module into Terahertz Time - domain Spectrum System （ THz - TDS）, set echo wave delay to 30 - 33 ps, which realized a THz tomography to interior structure of grain of wheat. Then, we adopted Maximum Between - Class Variance method （OTSU） to split vermiculate areas in obtained THz image. Last, we introduced Genetic Algorithm （GA） to improve the efficiency of OTSU in solving optimal threshold value. The experimental results show that the proposed method can recognize the vermiculate area in terahertz images of grains of wheat effectively.

关 键 词：蛀食性害虫 太赫兹成像 图像识别 最大类间方差法遗传算法 

分 类 号：O433.4[机械工程—光学工程]