基于形态特征的破碎玉米籽粒识别及检测装置设计

杨亮; 王卓; 白晓平

doi:10.13427/j.cnki.njyi.20240019.015

基于形态特征的破碎玉米籽粒识别及检测装置设计

杨亮^1,2,3,
王卓^1,2,3,,
白晓平^1,2,3

1. 中国科学院沈阳自动化研究所;
2. 中国科学院机器人与智能制造创新研究院;
3. 辽宁省农业装备智能化技术重点实验室

基金项目:

辽宁省重点研发计划项目(2020JH2/10200026)

国家重点研发计划项目(2020YFB1709600)

详细信息

作者简介:
杨亮(1993-),男,辽宁朝阳人,工程师,(E-mail)yangliang2@sia.cn

通讯作者:
王卓(1976-),男,沈阳人,研究员,博士生导师,(E-mail)zwang@sia.cn

中图分类号: TP391.41;S225.51
计量
- 文章访问数: 3
- HTML全文浏览量: 0
- PDF下载量: 0
出版历程
- 收稿日期: 2022-11-01
- 刊出日期: 2024-10-31

Recognition of the Broken Corn Kernels Based on Morphology Feature And Design of Detection Device

YANG Liang^1,2,3,
WANG Zhuo^1,2,3,,
BAI Xiao-ping^1,2,3

1. Shenyang Institute of Automation Chinese Academy of Sciences;
2. Institutes of Robotics and Intelligent Manufacturing, Chinese Academy of Sciences;
3. Key Laboratory of Agricultural Equipment Intelligent Technology in Liaoning Province

摘要

摘要: 基于K-Means聚类分割算法、颜色空间转换和形态学特征，通过对辽沈地区广泛种植的玉米的面积、面积与周长比、短轴与长轴比、圆形与矩形及五种形态特征的分析，识别出玉米碎粒。同时，设计开发了一种基于同步带轮的玉米籽粒破碎率在线检测装置，用于玉米籽粒破碎率的在线检测与识别。研究结果：(1)利用K平均算法分割算法将图像分割为目标区域和背景区域；(2)利用K平均算法分割算法将图像分割为二值化图像，进一步对二值化图像采用形态闭运算填充目标区域的非连通区域，对目标区域进行统计分析，计算出形态特征；(3)通过多组试验，玉米籽粒破碎的识别率可达94%。
- 玉米籽粒 /
- K-Means聚类分割 /
- 形态特征 /
- 颜色空间 /
- 在线检测装置
Abstract: Study on the K-Means clustering segmentation algorithm, color space conversion and morphology feature. Analyzing the corns’ area, the ratio of area and perimeter, the ratio of short axis and long axis, roundness and rectangularity, five morphology features to recognize the broken corn kernels that they are widely planted in Liao-Shen Area. At the same time, an online detection device for corn kernels broken rate based on synchronous pulley is designed and developed for online detection and recognition of corn kernels broken rate. Results:（1） using the K-Means clustering segmentation algorithm to divide the images into the target region and the background region;（2）The images have divided by K-Means clustering segmentation algorithm convert to the binarized images, further the binarized images use the morphological close operation to fill these are not connected areas of the target region, statistical analysis of the target region and we calculate the morphology features;（3）Through many groups of experiments, the recognition rate of corn kernels breakage can reach up to 94%.
- corn kernels /
- K-Means clustering segmentation /
- morphology feature /
- color space /
- online detection device

HTML全文

参考文献(23)

[1]	权龙哲,马小愚.基于小波分析的玉米籽粒图像正形研究[J].农机化研究,2006(2):154-156..
[2]	程洪,史智兴,尹辉娟,等.基于机器视觉的多个玉米籽粒胚部特征检测[J].农业工程学报,2013(19):145-151.
[3]	陈家娟,纪寿文,马成林,等.基于遗传神经网络的玉米叶色的自动研究测定[J].农业工程学报,2000,16(5):115-117.
[4]	史智兴,程洪,李江涛,等.图像处理识别玉米品种的特征参数研究[J].农业工程学报,2008(6):193-195.
[5]	DING K,GUNASEKARAN S.Shape feature extraction and classification of food material using computer vision[J].Trans ASAE,1994,37(5):1537-1545.
[6]	NI B,M R PAULSEN,J F REID.Corn kernel crown shape identification using image processing[J].Trans of the ASAE,1997,40(3):833-838.
[7]	NG H F,W F WICKLE,R V MOREY,et al.Machine vision evaluation of corn kernel mechanical and mold damage[J].Trans of the ASAE,1998,41(2):415-420.
[8]	ZAYAS I,H CONVERSE,J L STEELE.Discrimination of whole from broken corn kernels with image analysis[J].Trans of the ASAE,1990,33(5):1642-1646.
[9]	李晓峰,接鑫,张永丽,等.玉米种子内部机械裂纹检测与机理研究[J].农业机械学报,2010(12):143-147.
[10]	李心平,高春燕,刘赢,等.玉米果穗喂入形式与籽粒破碎率的关系研究[J].农机化研究,2013,35(12):137-140.
[11]	高国琴,李明.基于K-means算法的温室移动机器人导航路径识别[J].农业工程学报,2014,30(7):25-32.
[12]	尹建芹,刘小丽,田国会,等.基于关键点序列的人体动作识别[J].机器人,2016,38(2):200-207,216.
[13]	贾伟宽.基于智能优化的苹果采摘机器人目标识别研究[D].镇江:江苏大学,2016.
[14]	罗陆锋,邹湘军,王成琳,等.基于轮廓分析的双串叠贴葡萄目标识别方法[J].农业机械学报,2017,48(6):15-22.
[15]	BALDEBHAI P J,ANAND R S.Color image segmentation for medical images using L* a* b* color space[J].IOSR journal of electronics and communication engineering,2012,1(2):24-45
[16]	PASCHOS G.Perceptually uniform color spaces for color texture analysis:an empirical evaluation[J].IEEE transactions on image processing,2001,10(6):932-937.
[17]	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(14):202-208.
[18]	ABUALHAJ B,WENG G,ONG M,et al.Comparison of five cluster validity indices performance in brain [18F] FET-PET image segmentation using k-means[J].Medical physics,2017,44(1):209-220.
[19]	JAROŠ M,STRAKOŠ P,KARÁSEK T,et al.Implementation of K-means segmentation algorithm on intel xeon phi and GPU:Application in medical imaging[J].Advances in engineering software,2017,103:21-28.
[20]	霍迎秋,秦仁波,邢彩燕,等.基于CUDA的并行K-means聚类图像分割算法优化[J].农业机械学报,2014(11):47-53,74.
[21]	李震,洪添胜,曾祥业,等.基于K-means聚类的柑橘红蜘蛛图像目标识别[J].农业工程学报,2012(23):147-153,299.
[22]	杨锦忠,郝建平,杜天庆,等.基于种子图像处理的大数目玉米品种形态识别[J].作物学报,2008(6):1069-1073.
[23]	张玉荣,王伟宇,周显青,等.基于外观特征识别玉米不完善粒检测方法[J].河南工业大学学报(自然科学版),2015(2):1-7,12.