基于K-means与KNN的多特征苹果在线分级

doi:10.6048/j.issn.1001-4330.2023.03.015

新疆农业科学 ›› 2023, Vol. 60 ›› Issue (3): 643-650.DOI: 10.6048/j.issn.1001-4330.2023.03.015

• 植物保护·设施农业·农产品加工工程·微生物 • 上一篇下一篇

基于K-means与KNN的多特征苹果在线分级

王迎超(), 张婧婧(), 达新民, 耿新雪

新疆农业大学计算机与信息工程学院,乌鲁木齐 830052

收稿日期:2022-08-18 出版日期:2023-03-20 发布日期:2023-04-18
通信作者: 张婧婧(1981-),女,湖南宁乡人,副教授,硕士生导师,研究方向为农业信息化技术,(E-mail)zjj@xjau.edu.cn
作者简介:王迎超(1997-),男,河南驻马店人,硕士研究生,研究方向为图像处理,(E-mail)945697714@qq.com.
基金资助:
新疆维吾尔自治区自然科学基金(2022D01A202);新疆农业大学研究生教育教学改革研究项目(xjaualk-yjs-2021012)

Design of Multi-feature Apple Online Grading System Based on K-means and KNN

WANG Yingchao(), ZHANG Jingjing(), DA Xinmin, GENG Xinxue

College of Computer and Information Engineering, Xinjiang Agricultural University, Urumqi 830052, China

Received:2022-08-18 Online:2023-03-20 Published:2023-04-18
Correspondence author: ZHANG Jingjing (1981-), female, Hunan Province,associate professor, Research direction: agricultural information technology,(E-mail)zjj@xjau.edu.cn
Supported by:
Natural Science Foundation of Xinjiang Uygur Autonomous Region(2022D01A202);Postgraduate Education and Teaching Reform of Xinjiang Agricultural University(xjaualk-yjs-2021012)

摘要/Abstract

摘要：

【目的】准确获取红富士苹果的分级指标,为实现多特征融合的苹果分级提供依据。【方法】以均值滤波、全局亮度均衡化与图像裁剪方法,预处理实验所需的苹果图像;使用K-means聚类算法、OTSU最大类间方差法,将苹果灰度图转换为二值图;利用二值图与苹果原图的异或运算,提取苹果轮廓;采用苹果的二值图计算苹果的果实区域大小;使用颜色空间转换RGB-HSV中H通道划分果实红色区域;通过构建掩膜、形态学操作判断果体是否含有缺陷及计算其面积;构建最小外接矩形计算苹果的果径及果形;利用KNN分类算法实现多特征融合的苹果在线自动分级。【结果】基于K-means聚类与KNN分类相结合的苹果在线分级方法,在优于传统图像阈值分割效果的基础上,特级果分级准确率为97.14%,一级果分级准确率为100%,二级果分级准确率为93.75%,等外果分级准确率为100%,综合分级准确率达到97%。【结论】100个苹果测试准确率达到97%,验证了该分级方法的可行性与准确性。

关键词: 多特征融合; 苹果自动分级; K-means; OTSU; KNN

Abstract:

【Objective】 To study the automatic classification method based on the combination of K-means clustering and KNN nearest neighbor classification with a view to providing basis for accurately obtaining the classification index of Red Fuji apple and realizing multi feature fusion apple classification.【Methods】 The apple images needed in the experiment were preprocessed by means of mean filtering, global brightness equalization and image clipping; K-means clustering algorithm and OTSU maximum interclass variance method were used to convert apple gray image into binary image; Using the XOR operation between the binary graph and the original apple graph, the apple contour is extracted; The area size of apple fruit was calculated by using the binary map of apple; The H-channel in RGB-HSV was transformed into red region of fruit by color space; The construction of mask and morphological operation was applied to judge whether the fruit body contained defects and calculate its area; The minimum circumscribed rectangle was constructed to calculate the fruit diameter and shape of apple; KNN classification algorithm was used to realize on-line automatic classification of apple with multi feature fusion. 【Results】 The experimental results show that the automatic Apple classification system based on K-means clustering and KNN classification has a classification accuracy of 97% without reducing the effect of traditional image threshold segmentation. 【Conclusion】 Through testing 100 apples,the accuracy reached 97%,which verified the feasibility and accuracy of the grading method.

Key words: multi feature fusion; automatic apple grading; K-means; OTSU; KNN

中图分类号:

王迎超, 张婧婧, 达新民, 耿新雪. 基于K-means与KNN的多特征苹果在线分级[J]. 新疆农业科学, 2023, 60(3): 643-650.

WANG Yingchao, ZHANG Jingjing, DA Xinmin, GENG Xinxue. Design of Multi-feature Apple Online Grading System Based on K-means and KNN[J]. Xinjiang Agricultural Sciences, 2023, 60(3): 643-650.

图/表 16

图1 苹果分级

Fig.1 Apple grading principle

图2 苹果亮度均衡化前后

Fig.2 Display of Apple brightness before and after equalization

图3 苹果背景分离

Fig.3 Apple background separation

图4 颜色特征提取

Fig.4 Color feature extraction

表1 HSV基本颜色分量范围

Tab.1 The HSV basic color component range

	黑 Black	灰 Grey	白 White	红 Red		橙 Orange	黄 Yellow	绿 Green	青 Young	蓝 Blue	紫 Violet
H_min	0	0	0	0	156	11	26	35	78	100	125
H_max	180	180	180	10	180	25	34	77	99	124	155
S_min	0	0	0	43		43	43	43	43	43	43
S_max	255	43	30	255		255	255	255	255	255	255
V_min	0	46	221	46		46	46	46	46	46	46
V_max	46	220	255	255		255	255	255	255	255	255

图5 苹果果径测量

Fig.5 Apple fruit diameter measurement

图6 苹果缺陷检测

Fig.6 Apple defect detection

图7 红富士苹果在线分级系统界面设计

Fig.7 Interface design of red fuji apple online grading system

图8 预处理效果

Fig. 8 Pretreatment effect

图9 苹果图像背景分离

Fig. 9 Apple image background separation

图10 红色特征提取

Fig. 10 Red feature extraction

图11 果径测量

Fig. 11 Effect of fruit diameter measurement test

图12 苹果缺陷提取

Fig. 12 Apple defect extraction

表2 分级比较

Tab.3 Grading results

	真实值(个) True value	预测值(个) Estimate	误差(个) Error
特级果 Super fruit	35	34	1
一级果 Primary fruit	24	24	0
二级果 Secondary fruit	32	30	2
等外果 Epicarp	9	9	0

表3 各等级苹果的分析结果

Tab.4 Grading rdsults of apple of different grades

图13 各分级指标

Fig. 13 Classification indexes

参考文献 13

[1]	Yu Y, Velastin S A, Yin F. Automatic apple grading based on multi feature and weighted K-means Clustering algorithm[J]. Agricultural Information Processing, 2020, 7(4):10.
[2]	Hu Z L, Tang J S, Zhang P, et al. Deep learning for the identification of bruised apples by fusing 3D deep features for apple grading systems[J]. Mechanical Systems and Signal Processing, 2020:106922.
[3]	Mhapne N V, Harish S V, Kini A S, et al. A Comparative Study to find an Effective Image Segmentation Technique using Clustering to obtain the Defective Portion of an Apple[C]. 2019 International Conference on Automation, Computational and Technology Management (ICACTM). 2019.
[4]	廖崴, 郑立华, 李民赞, 等. 基于随机森林算法的自然光照条件下绿色苹果识别[J]. 农业机械学报, 2017, 48(增刊):86-91.
	LIAO Wei, ZHENG Lihua, LI Minzan, et al. Green apple recognition under natural illumination based on random forest algorithm[J]. Journal of Agricultural Machinery, 2017, 48(supplement): 86-91.
[5]	张保华, 黄文倩, 李江波, 等. 基于亮度校正和AdaBoost的苹果缺陷在线识别[J]. 农业机械学报, 2014, 45(6):221-226.
	ZHANG Baohua, HUANG Wenqian, LI Jiangbo, et al. Apple defect online recognition based on brightness correction and AdaBoost[J]. Journal of Agricultural Machinery, 2014, 45(6):221-226.
[6]	Bhatt A K, Pant D. Automatic apple grading model development based on back propagation neural network and machine vision,and its performance evaluation[J]. Ai& Society, 2013, 30(1):45-56.
[7]	刘云, 杨建滨, 王传旭. 基于卷积神经网络的苹果缺陷检测算法[J]. 电子测量技术, 2017, 40(3):108-112.
	LIU Yun, YANG Jianbin, WANG Chuanxu. Apple defect detection algorithm based on convolutional neural network[J]. Electronic Measurement Technology, 2017, 40(3): 108-112.
[8]	赵娟, 彭彦昆, SAGAR Dhakal, 等. 基于机器视觉的苹果外观缺陷在线检测[J]. 农业机械学报, 2013, 44(增刊):260-263.
	ZHAO Juan, PENG Yankun, Sagar Dhakal, et al. Online detection of Apple appearance defects based on machine vision[J]. Journal of Agricultural Machinery, 2013, 44(supplement): 260-263
[9]	Yang Y A, Sav C, Fei Y. Automatic grading of apples based on multi-features and weighted K-means clustering algorithm[J]. Information Processing in Agriculture, 2019.
[10]	杨刚, 张先林, 张江. 果形素对苹果果形指数影响的研究[J]. 林业科技开发, 2001,(1):28-29.
	YANG Gang, ZHANG Xianlin, ZHANG Jiang. Study on the effect of fruit shape element on apple fruit shape index[J]. Forestry Science and Technology Development, 2001,(1):28-29.
[11]	辛淑亮, 王奎先, 林振海, 等. 采收期苹果、梨果实纵横径与果形指数的理论分析[J]. 莱阳农学院学报, 1986,(1):63-71.
	XIN Shuliang, WANG kuixian, LIN Zhenhai, et al. Theoretical analysis of fruit vertical and horizontal diameter and fruit shape index of apple and pear at harvest time[J]. Journal of Laiyang Agricultural College, 1986,(1):63-71.
[12]	Zhao D, Hu X, Xiong S, et al. K-means clustering and kNN classification based on negative databases[J]. Applied Soft Computing 2021, 110(1):107732. DOI URL
[13]	Liu C, Wang W, Wang M, et al. An efficient instance selection algorithm to reconstruct training set for support vector machine[J]. Knowledge-Based Systems, 2017, 116(1): 58-73.

基于K-means与KNN的多特征苹果在线分级

Design of Multi-feature Apple Online Grading System Based on K-means and KNN

在线阅读

PDF下载

可视化

摘要/Abstract

引用本文

使用本文

图/表 16

参考文献 13

相关文章 1

编辑推荐

Metrics