Simulink平台在玉米病害视频图像中的实时诊断  

Simulink Platform in Video Image Real-time Diagnosis of Maize Disease

作　　者：孟颖[1] 陈桂芬[1] 卢建[1] 许荣泽 

机构地区：[1]吉林农业大学信息技术学院,长春130118

出　　处：《吉林农业大学学报》2017年第4期483-487,共5页Journal of Jilin Agricultural University

基　　金：国家"863"项目(2006AA10A309);国家星火计划项目(2008GA661003);吉林省财政厅世界银行贷款项目(2011-Z20);吉林省农委项目(2015)

摘　　要：为提高对玉米病害叶片的识别精度,达到快速诊断、智能决策和有效诊治的目的,提出了一种基于Simulink仿真平台的玉米病害视频图像远程实时诊断技术。该技术首先使用Simulink仿真平台将采集的实时视频进行平滑处理,以提高图像的清晰度和质量;再运用分割技术确定玉米病害的优选图像;最后进行优选图像的解析和诊断处理。对玉米灰斑病图像的研究结果表明:该优化技术处理后的图像质量明显提高,突出了玉米病害特征,增强了玉米病害远程视频图像诊断的实时性和准确性,为玉米生产的智能决策提供了技术支撑。To improve recognition accuracy of diseased maize leaves and to achieve the aims of rapid diagnosing,intelligent decision making and effective diagnosing and treating,this study proposed a remote and real-time video image diagnosis technology of maize disease based on Simulink simulation platform. Simulink simulation platform was firstly used to process real-time video data captured,including smoothing,to improve image clarity and quality; Secondly,segmenting was used to determine the optimal image of maize diseases; Finally,the optimal images were analyzed and diagnosed.The study results of gray leaf spot of maize images show that the quality of images after optimal algorithm improved remarkably,the characteristics of maize diseases were highly lightened,and realtime performance and accuracy of the diagnosis on video image recognition technology of maize disease were enhanced,which will provide technical support for intelligent decision making of maize production.

关 键 词：玉米灰斑病 实时视频图像 simulink仿真平台 优选图像 

分 类 号：S126[农业科学—农业基础科学] S435.131