【作者】 李睿远；

【导师】 柴苍修；

【作者基本信息】 武汉理工大学 ， 机械电子工程， 2007， 硕士

【摘要】 目前我国农林业病虫害形势严峻，而防治效率很低，主要是由于我国的植保机械落后和试药技术低下造成的。为了适应农林业的发展要求，本文在参阅了国内外喷雾机先进技术的基础上，将电子，信息和自动化控制等技术和低容量喷雾、离心雾化、风送远射程喷雾技术引入到喷雾机，研制了风送式超低容量喷雾机。本文首先对超低容量喷雾机的关键系统一风路系统的喷口作了设计，针对对五种喷口形状做了风速和距离的关系试验，然后选取合理的喷口形状，并根据最小二乘法拟合了公式，并且从理论上验证了所拟合的公式的正确性，这对以后喷筒的设计提供了基本的理论依据。其次设计了液位的自动检测系统和光声报警装置，由于喷雾机的药箱都是内置的，其剩余药液的多少我们很难很快知道，引入了液位的自动检测和显示模块后就可以方便的知道实时液位情况，并且在药液很少时它可以报警提醒试药人员，这样不但提高了超低容量喷雾机的自动化程度，而且还减轻了试药人员的劳动强度，提高了工作效率。第三，深入探讨了数字图像处理的理论，对图像的多种预处理方法和多种边缘检测分割方法作了对比研究。在将原始彩色图像转变为灰度图像后，进行了直方图均衡化处理和滤波处理，增强图像的对比度。在预处理后，采用自动阈值分割法分割出了目标靶的轮廓，并且提取出目标靶的关键参数，如形心坐标，面积等。最后在对图像分割结果的基础之上，然后根据图像形心在图像中的位置关系将图像进行分类，并根据喷嘴的喷雾参数和摄像机的几个关键参数，对各个喷嘴分别进行控制，包括控制每个喷嘴的开闭与否以及控制喷嘴的步进摆动等。更多还原

【Abstract】 Presently the infestation of the forest and agriculture plant diseases and insects inChina causes severe problems, while the efficiency of treatment is very low, Becausethe plant-protection equipments are poor, and the spray technology of our country islag. In order to meet the developing requirement of the forest and agriculture, thispaper brought electronics, information and automation technology andultra-low-volume spraying, centrifugal spraying, air-assisted spraying technologyinto sprayers, and finally manufactured air-assisted ultra-volume sprayer.First, we designed the spout’s structure of air-pipe system which was theultra-volume sprayer’s crucial system, and we made experiments on five differentspouts about the relation between the wind velocity and the spraying distance, thenselected the reasonable structure, and finally fit a formulary based on the least squaremethod.Second, we designed the liquid-level measure device and the light-sound alarmannunciator, because the tank of the sprayer usually was placed inside the cover, sowe cannot know the real-time liquid-lever. After designing the liquid-level automeasure device, the light-sound alarm annunciator and displaying module, we canknow conveniently the real-time liquid-level of the tank. and what’s more, when theliquid medicine is under the threshod which is set first, it can give an alarm.Third, after studying the theory of the digital image processing, we contrastedseveral preprocessing methods, several edge detection and segmentation methods.the first step, changing the original color image to gray image, processing the grayimage by histogram equalization and filter operation to enhance the image’s contrast.The second step, after preprocessing, the profile of the target image was segmentedout using auto threshoding segmentation method, the last step, figureing out the keyparameters, eg: area, centroid.Finally, based on the results of the image processing, the images wereclassified, and each nozzle was controlled by the single chip.更多还原