基于机算机视觉的杨梅自动检测分级
|
摘要
我国是杨梅的发源地和主产区,地理环境差异大,形成了多样的品种和品质。杨梅以鲜果形式进行销售和食用,销售期很短,因此快速高效的分级具有重要意义。人工分级的劳动强度大、效率低且分级质量不稳定,这些问题在杨梅上体现得更为明显。计算机视觉技术是水果分级研究中的热点,目前已经安装在杨梅采摘机器人上,用于自然环境下的果实识别。本文设计了基于计算机视觉的杨梅自动检测分级系统,经过计算机视觉软件的预处理、灰度化和图像分割后提取杨梅轮廓;然后检测鲜果果径、圆形度和颜色深度,根据相应的等级标准进行分级。系统分级的准确达到92.7%,对大小和果形均具有较高的识别准确率;单幅图片处理耗时0.45s,能够满足实时检测分级的要求。这个系统与合适的分级执行装置结合使用可以达到高效无损的分级效果,为我国杨梅产业可持续发展提供技术支撑。
Yangmei nutrient-rich,with high food and medicinal value.China is the birthplace of Yangmei and the main producing areas,geographical differences,forming a variety of varieties and quality.Yangmei in the form of fresh fruit for sale and consumption,the sales period is very short,fast and efficient classification is of great significance.Artificial grading of labor intensity,low efficiency and poor quality grading,these problems in the Yangmei reflected more obvious.Computer vision technology is a hot spot in fruit grading research,has been installed in the bayberry picking robot,for the natural environment of fruit recognition.In this paper,the automatic detection grading system of Yangmei based on computer vision is designed,and the contours of Yangmei are extracted by computer vision software pretreatment,gray scale and image segmentation.And then detect fruit fruit diameter,roundness and color depth,according to the corresponding level of classification.The system classification to achieve accurate 92.7%,the size and shape of a high recognition accuracy.Single image processing takes 0.45 s,which can meet the requirements of real-time detection grading.This system can be used in combination with the appropriate grading execution.
Yangmei nutrient-rich,with high food and medicinal value.China is the birthplace of Yangmei and the main producing areas,geographical differences,forming a variety of varieties and quality.Yangmei in the form of fresh fruit for sale and consumption,the sales period is very short,fast and efficient classification is of great significance.Artificial grading of labor intensity,low efficiency and poor quality grading,these problems in the Yangmei reflected more obvious.Computer vision technology is a hot spot in fruit grading research,has been installed in the bayberry picking robot,for the natural environment of fruit recognition.In this paper,the automatic detection grading system of Yangmei based on computer vision is designed,and the contours of Yangmei are extracted by computer vision software pretreatment,gray scale and image segmentation.And then detect fruit fruit diameter,roundness and color depth,according to the corresponding level of classification.The system classification to achieve accurate 92.7%,the size and shape of a high recognition accuracy.Single image processing takes 0.45 s,which can meet the requirements of real-time detection grading.This system can be used in combination with the appropriate grading execution.
引文
[1]何新华,陈力耕,陈怡,等.中国杨梅资源及利用研究评述[J].果树学报,2004,21(5):467-471.
[2]陈方永.我国杨梅研究现状与发展趋势[J].中国南方果树,2012,41(5):31-36.
[3]梁森苗,王勤红,倪国富,等.东魁杨梅品质栽培关键生产技术概述[J].现代园艺,2012(1):32-33.
[4]曹雪丹,戚行江,方修贵,等.果实干燥特性及其品质[J].浙江农业学报,2010,22(3):366-369.
[5]郑丽屏,蔡平,王利芬,等.苏州地区8个杨梅品种果实品质分析[J].江苏农业科学,2013,41(7):159-160.
[6]李莉,林文欢,郑道序,等.汕头市6个杨梅品种的特性与评价[J].湖北农业科学,2016,55(19):5504-5506.
[7]袁启凤,范建新,张诗莹、等.贵州杨梅地方品种果实的品质分析[J].贵州农业科学,2010,38(10):171-172.
[8]龚洁强,黄茜斌,林媚.东魁杨梅果实大小与品质的相关分析[J].浙江柑橘,2007,24(2):31-32.
[9]黄海静,符国槐,杨再强,等.设施栽培对杨梅生长发育和品质的影响[J].南京林业大学学报:自然科学版,2012,36(6):47-52.
[10]蒋玲玲.颜色空间结合化学计量在杨梅采后筛选及加工储藏中的应用研究[D].杭州:浙江师范大学,2014.
[11]蒋巧俊,徐静.采后商品化处理技术研究进展[J].保鲜与加工,2011,11(4):30-34.
[12]徐黎明,吕继东.基于同态滤波和K均值聚类算法的杨梅图像分割[J].农业工程学报,2015,31(14):202-208.
[13]徐黎明,吕继东.自然环境下杨梅果实图像的分割方法研究[J].沈阳农业大学学报,2016,47(3):334-341.
[2]陈方永.我国杨梅研究现状与发展趋势[J].中国南方果树,2012,41(5):31-36.
[3]梁森苗,王勤红,倪国富,等.东魁杨梅品质栽培关键生产技术概述[J].现代园艺,2012(1):32-33.
[4]曹雪丹,戚行江,方修贵,等.果实干燥特性及其品质[J].浙江农业学报,2010,22(3):366-369.
[5]郑丽屏,蔡平,王利芬,等.苏州地区8个杨梅品种果实品质分析[J].江苏农业科学,2013,41(7):159-160.
[6]李莉,林文欢,郑道序,等.汕头市6个杨梅品种的特性与评价[J].湖北农业科学,2016,55(19):5504-5506.
[7]袁启凤,范建新,张诗莹、等.贵州杨梅地方品种果实的品质分析[J].贵州农业科学,2010,38(10):171-172.
[8]龚洁强,黄茜斌,林媚.东魁杨梅果实大小与品质的相关分析[J].浙江柑橘,2007,24(2):31-32.
[9]黄海静,符国槐,杨再强,等.设施栽培对杨梅生长发育和品质的影响[J].南京林业大学学报:自然科学版,2012,36(6):47-52.
[10]蒋玲玲.颜色空间结合化学计量在杨梅采后筛选及加工储藏中的应用研究[D].杭州:浙江师范大学,2014.
[11]蒋巧俊,徐静.采后商品化处理技术研究进展[J].保鲜与加工,2011,11(4):30-34.
[12]徐黎明,吕继东.基于同态滤波和K均值聚类算法的杨梅图像分割[J].农业工程学报,2015,31(14):202-208.
[13]徐黎明,吕继东.自然环境下杨梅果实图像的分割方法研究[J].沈阳农业大学学报,2016,47(3):334-341.