作物病害机器视觉诊断研究进展  

Advances in Research on Computer-Vision Diagnosis of Crop Diseases

作　　者：赖军臣[1,2] 李少昆[1,2] 明博[1,2] 王娜[1,2] 王克如[1,2] 谢瑞芝[2] 高世菊[2] 

机构地区：[1]新疆兵团绿洲生态农业重点开放实验室/新疆作物高产研究中心,新疆石河子832003 [2]中国农业科学院作物科学研究所/国家农作物基因资源与基因改良重大科学工程,北京100081

出　　处：《中国农业科学》2009年第4期1215-1221,共7页Scientia Agricultura Sinica

基　　金：国家“863”项目(2007AA10Z237、2006AA10Z207)

摘　　要：据统计,由病害引起的作物产量损失平均在12%以上,病害不仅直接影响产量,而且也严重降低农产品的品质。通过快速诊断识别作物病害而及时采取防治措施,提高作物病害的诊断水平是保证有效防治病害的重要组成部分。机器视觉技术的发展为作物病害诊断和识别提供了快捷、低廉、无损检测的可能手段。本文综述了面向多种作物病害机器视觉诊断研究的进展,归纳了作物病害机器视觉诊断的识别模式,分析了诊断中存在的问题,提出了相应的解决方法。Statistics indicate that the crops output average loss caused by plant disease is above 12%, plant diseases not only affect the output directly, but also seriously reduce the quality of agricultural products. Taking the preventing and controlling measures promptly through rapid diagnosis and identification of crop diseases and improving the diagnosis level of crop diseases are an important part in preventing and controlling plant disease effectively. Development of computer-vision technology makes it possible to diagnose crop diseases quickly, economically and no-destructively. Firstly, the new progress of computer-vision diagnoses of crop diseases is reviewed, and the four types of technology routines of diagnosis of crop diseases are summarized. Secondly, the difficulties of diagnosing crop diseases with computer-vision are analyzed, and the possible solutions are discussed. Finally the development trend of computer-vision diagnosis of crop diseases is discussed.

关 键 词：作物 病害 机器视觉 诊断 进展 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]