利用图像处理技术检测粮仓害虫的研究

摘要

运用图像处理技术实现粮仓害虫的检测．通过特制的粮食取样器在粮仓中采集图像序列，实时地进行图像增强和二值化处理，并在线给出有无害虫的图像、文字和声警信息，利用切换采样点时间进行图像特征的提取、分类等深层信息处理，从而还可以给出有关粮仓害虫种类、密度和位置等相关信息，并结合现有成果和专家经验给出相应的操作指导建议．文中给出了系统的总体设计及其实现方案．研究结果表明，该系统获得了较好的检测效果，这一思路为粮仓害虫的检测与识别提出了一种新方法．%The image processing technology to detect insects in grainelevator is tried to use.By special designed foodstuff sampler,the CCD col lect,intensify and threshold the image series in bin.The system will give prompt and warning information and obtain the result that whether the sample site has insects or not.Furthermore,the system use the sampling time interval to draw out the character and classify in order to offer the category,density and position information of the insects,meanwhile it will give corresponding measure which th e worker should do combine with the expert’s experiences.It also shows the tota l design and execution scheme of the detection system.The result of research ind icates that system gains a quite good detecting effectiveness and this method br ings about a new thinking for detecting grain elevator insects.