基于图像处理技术番茄植物叶片检测方法的研究.pdfVIP

基于图像处理技术番茄植物叶片检测方法研究 姜伟杰，孙明，赵卓英 【摘要】本文以温室条件下的番茄叶片为研究对象，研究了适合于番茄叶片特征检测的图像处理方法， 为虚拟植物建模及其可视化提供精确可靠的数据支持。首先。用数字图像采集技术进行图像的采集，分析 产生失真的机理，并成功地对透视失真的番茄叶片图像进行了校正。然后，研究了叶片图像的点运算、增 强技术、边缘检测及轮廓跟踪等算法，实现了叶片几何特征的计算。通过对比分析，得到了比人工测量更 高精度的结果，表明基于数字图像处理的作物生长检测方法无接触、低费用、精度高，是一种切实可行的 检测方法。 关键词图像处理：番茄叶片；特征提取 Researchof tomatobasedon detectingleaves imageprocessing Weijie ZHAO JIANG,MingSUN，Zhuoying (China AgriculturalUniversity) AbstractTomatoleaves were aS in chosenresearch inthe method greenhouse objectspaper．Theimageprocessing wasdiscussedfor theirfeaturesinorderto data virtual tomatoes for of detecting provide plant．Firstly’theimages werecollected the of of distortionWaS by technology using digitalimagegrabbing．neprinciplegeometric thenthe correctionmethodsuitabletothe distortionof tomatoleafWaSobtained． analyzed，and perspective every studiedthe suchas contour Secondly,wealgorithmsgray-leveltransformation，imagefiltering，imagetracking，etc． was Featureextraction totraditionalman．made moreaccurateresultWaS accomplished．Compared measurement,a inthis themethod of tomatoleavesbasedon acquiredpaper．Itproved detecting non—destructive, imagesprocessing feasible． lOW—cost’moreaccurate，and words extraction Key