A new proposal for automatic identification of multiple soybean diseases

被引：27

作者：

Macedo Araujo, Juliana Mariana ^{[1
]}

Assis Peixoto, Zelia Myriam ^{[1
]}

机构：

[1] Pontificia Univ Catolica Minas Gerais, Grad Program Elect Engn, Av Itau 525, BR-30535012 Belo Horizonte, MG, Brazil

来源：

COMPUTERS AND ELECTRONICS IN AGRICULTURE | 2019年 / 167卷

关键词：

Soybean leaf disease; Local binary pattern; Support vector machine; Color moments; Bag of visual words model; CLASSIFICATION; PLANT; CROPS; WEEDS;

D O I：

10.1016/j.compag.2019.105060

中图分类号：

S [农业科学];

学科分类号：

09 ;

摘要：

This work proposes a combination of digital image processing techniques consisting of color moments, local binary patterns (LBP) and bag of visual words (BoVW) model for automatic detection of soybean diseases based on the analysis of color, texture and local characteristics of spots on affected leaves. The characteristics extracted after data collection and the application of the cited techniques are grouped and used as input to the support vector machine (SVM) for the purpose of disease classification. The images used in the development and validation of the proposed identification system were obtained from Digipathos, a database provided by the Brazilian Agricultural Research Agency (Embrapa), consisting of 354 images related to 8 typical soybean diseases, namely, bacterial blight, soybean rust, copper phytotoxicity, soybean mosaic, target spot, downy mildew, powdery mildew and Septoria brown spot. Performance analyses are presented separately for the techniques used in several stages of data processing and for the overall system consisting of the combination of these techniques and the SVM classifier. The proposed system achieves an success rate of 75.8%, representing an improvement of approximately 17% over results of methods in the existing literature.

引用

页数：9

共 46 条

[11]

Cherkassky V, 1997, IEEE Trans Neural Netw, V8, P1564, DOI 10.1109/TNN.1997.641482

[12]

d Albuquerque R.W., 2012, THESIS

[13]

do Amaral V., 2013, P 10 ENC NAC INT ART

[14]

Evans C., 2009, NOTES OPENSURF LIB

[15]

Fiel Stefan., 2010, Automated identification of tree species from images of the bark, leaves and needles

[16] Support Vector Machines for crop/weeds identification in maize fields [J].

Guerrero, J. M. ;

Pajares, G. ;

Montalvo, M. ;

Romeo, J. ;

Guijarro, M. .

EXPERT SYSTEMS WITH APPLICATIONS, 2012, 39 (12) :11149-11155

[17] Kernel methods in machine learning [J].

Hofmann, Thomas ;

Schoelkopf, Bernhard ;

Smola, Alexander J. .

ANNALS OF STATISTICS, 2008, 36 (03) :1171-1220

[18] Evaluation of digital photography from model aircraft for remote sensing of crop biomass and nitrogen status [J].

Hunt Jr. E.R. ;

Cavigelli M. ;

Daughtry C.S.T. ;

McMurtrey III J.E. ;

Walthall C.L. .

Precision Agriculture, 2005, 6 (4) :359-378

[19] Dendroscope: An interactive viewer for large phylogenetic trees [J].

Huson, Daniel H. ;

Richter, Daniel C. ;

Rausch, Christian ;

Dezulian, Tobias ;

Franz, Markus ;

Rupp, Regula .

BMC BIOINFORMATICS, 2007, 8

[20]

Jian Z, 2010, 2ND IEEE INTERNATIONAL CONFERENCE ON ADVANCED COMPUTER CONTROL (ICACC 2010), VOL. 5, P264, DOI 10.1109/ICACC.2010.5487242

← 1 2 3 4 5 →