基于Android的棉花红蜘蛛虫害检测技术研究

发布时间：2018-05-31 11:16

  本文选题：棉花 + 红蜘蛛　； 参考：《石河子大学》2016年硕士论文

【摘要】：棉花一直以来是我国农业的重要经济作物,棉花种植业作为农业的支柱性产业之一,促进了我国纺织工业的发展和国民经济的提升。棉花的病虫害影响了棉花的产量和质量,根据统计每年因为棉花病虫害造成的产量损失约15%~20%,严重时能达到50%的损失,这大大影响了农民的收入和棉花产业的发展。传统的棉花病虫害诊断的方法主要是人工依据经验进行判断,这样主观性强,误差较大,棉农不能及时应对进行诊治,导致了棉花的减产。因此本文分析了当前的病虫害的检测技术手段,结合图像处理与Android系统完成了基于Android的棉花红蜘蛛虫害的检测系统的设计。棉花红蜘蛛危害严重时可以造成棉田减产30%以上,本文针对此虫害设计了一种基于Android的棉花红蜘蛛虫害检测系统,实时性好、应用性强、便携性好,为了实现此检测系统,主要对系统的结构功能、系统流程、软件设计、检测方法进行了研究,主要工作如下:(1)分析了基于Android的棉花红蜘蛛虫害检测系统的总体结构和功能需求,根据系统的设计原则选择移动终端平台,完成了对基于Android的棉花红蜘蛛虫害检测系统的总体和诊断流程的设计。(2)选择OpenCV机器视觉库,对获取的棉花红蜘蛛虫害的图像进行预处理、特征提取、识别和诊断,图像处理包括图像灰度化、图像分割、图像平滑和图像压缩等,特征提取包括对棉叶上受到红蜘蛛虫害侵害的部分进行了边缘、形状和颜色的提取,之后根据形状和颜色特征进行棉花红蜘蛛虫害的识别,最后根据像素比值法进行了棉花红蜘蛛受害程度的诊断。(3)配置完成Android系统通用开发环境的建立,利用Android应用的基本组件和应用资源完成了用户界面的设计,棉花红蜘蛛虫害图像的采集,功能主界面的浏览和使用,棉花红蜘蛛虫害相关资料查询和系统使用的帮助指南。(4)服务器端完成账号的基本操作,利用OpenCV机器视觉库的Java接口调用图像处理算法,将棉花红蜘蛛虫害处理、识别和检测的结果显示在手机移动终端上,完成基于Android的棉花红蜘蛛虫害检测系统应用程序,通过该程序进行测试,实现了棉花红蜘蛛虫害的检测功能。
[Abstract]:Cotton has always been an important economic crop in agriculture in China. As one of the pillar industries of agriculture, cotton planting has promoted the development of textile industry and the promotion of national economy. Cotton diseases and insect pests affect the yield and quality of cotton. According to statistics, the yield loss caused by cotton diseases and insect pests is about 15 / 20 per year, which can reach 50% loss when serious, which greatly affects the income of farmers and the development of cotton industry. The traditional diagnosis method of cotton diseases and insect pests is mainly based on experience, so the subjectivity is strong, the error is large, cotton farmers can not deal with timely diagnosis and treatment, resulting in cotton production. Therefore, this paper analyzed the current detection technology of pests and diseases, and completed the design of cotton red spider pest detection system based on Android combined with image processing and Android system. Cotton red spider can reduce the yield of cotton field by more than 30% when the harm is serious. This paper designs a cotton red spider pest detection system based on Android, which has good real-time performance, strong application, good portability, in order to realize this detection system. This paper mainly studies the structure function, system flow, software design and detection method of the system. The main work is as follows: 1) analyze the overall structure and function requirement of cotton red spider pest detection system based on Android. According to the design principle of the system, the mobile terminal platform is selected, and the OpenCV machine vision library is selected to design the whole system and diagnosis flow of cotton red spider pest detection system based on Android. Preprocessing, feature extraction, recognition and diagnosis of cotton red spider pest, image processing including image grayscale, image segmentation, image smoothing and image compression, etc. Feature extraction includes the extraction of edges, shapes and colors of parts of cotton leaves affected by red spider pests, and then identification of cotton red spider pests according to shape and color features. Finally, according to the pixel ratio method, the paper makes a diagnosis of the damage degree of cotton red spider. The configuration completes the establishment of the general development environment of the Android system, and uses the basic components and application resources of the Android application to complete the design of the user interface. Cotton red spider pest image collection, function interface browsing and use, cotton red spider pest related information query and system use help guide. 4) the server end to complete the basic operation of the account, Using the Java interface of OpenCV machine vision library to call the image processing algorithm, the result of cotton red spider pest processing, recognition and detection is displayed on mobile terminal of mobile phone, and the application program of cotton red spider pest detection system based on Android is completed. Through the test program, the detection function of cotton red spider pest was realized.
【学位授予单位】：石河子大学
【学位级别】：硕士
【学位授予年份】：2016
【分类号】：S435.622.2;TP391.41
，

本文编号：1959431