基于几何和色度学特征识别的花生视觉选种方法  

Visual seed selection method for peanuts based on geometric and chromaticity feature recognition

作　　者：王雪晴 郑福来 赵换丽 Wang Xueqing;Zheng Fulai;Zhao Huanli(Institute of Electrical and Mechanical Engineering,Pingdingshan University,Pingdingshan,467000,China)

机构地区：[1]平顶山学院电气与机械工程学院,河南平顶山467000

出　　处：《中国农机化学报》2021年第5期94-99,共6页Journal of Chinese Agricultural Mechanization

基　　金：河南省教育厅项目(17A413009);平顶山学院青年基金项目(PXY—QNJJ—2018001)。

摘　　要：花生选种要求花生粒完好、个大、颗粒饱满。为减少花生选种的工作量,该研究以花生粒为研究对象,建立花生选种的视觉识别方法。该方法首先采集图像并对图像进行视觉处理,根据训练样本图像建立基于几何和色度学的项目特征库;再根据采集的待选种子花生图像,对比特征库测试待选种子花生的2个重要几何特征参数面积和圆度,来识别粒大饱满种子花生;以及测试其色度学特征参数,区分完好、破损和霉变花生,从而识别能够选做种子的花生。通过识别测试,该研究的花生视觉选种识别率达到93.3%,能够满足选种要求,该花生视觉选种方法可应用于花生自动选种装置。Peanut seed is required to be intact,large,and plump.In order to reduce the workload of peanut seed selection,a visual recognition method for peanut mechanical seed selection is established.This method collects peanut seed images and visually processes the image,then tests two important geometrical parameters area and roundness of peanut to identify large and full seed peanut.The characteristic color parameters of the seeds are tested to distinguish between intact,damaged,and mildew peanuts so that the peanuts that can be selected as seeds are identified.The recognition rate of visual peanut selection is 93.3%,which meets the requirements for seed selection and can be applied to peanut automatic seed selection devices.

关 键 词：花生 选种 视觉识别 色度学特征 几何特征 图像处理 

分 类 号：S24[农业科学—农业电气化与自动化]