Techniques for Rice Leaf Disease Detection using Machine LearningAlgorithms

DOI : 10.17577/IJERTCONV9IS08032

Download Full-Text PDF Cite this Publication

Text Only Version

Techniques for Rice Leaf Disease Detection using Machine LearningAlgorithms

Dr. Sandhya Venu Vasantha

IT Dept.

M.V.S.R. Engineering College Hyderabad, India

Dr. Bejjam Kiranmai

CSE Dept.

KMIT, Hyderabad, India.

Dr. S. Rama Krishna

CSE Dept.

Bapatla Engineering College, Bapatla, India

Abstract Rice (Oryza Sativa) is a kind of cereal grain that is consumed as staple food by almost half of the worlds population, more specifically in Africa and Asia. The Crops of Oryza Sativa are exposed to both abiotic stresses such as cold, drought, salinity, etc., and biotic stresses such as insects, pests, bacterial, viral and fungal diseases. Furthermore it has become most challenging task for the farmer to identify the kind of disease the crop has affected with and which in fact affects the yield of the crop if not timely detected. This paper provides the possible solutions using various machine learning techniques and the comparative analysis of algorithms diagnosing the type of disease which has affected the crop based on the crops image data and more over it presents recently presented techniques with their performance measure.

Some of the significant diseases affecting the O. Sativa crop are detailed as follows:

  1. Brown Spot(BS)

    A fungal disease which infects the entire crop that can be easily identified in the early stages as it appears on the initial seedling leaves like brown oval or circular spots. The reason behind this is Bipolaris Oryzae a type of fungi, which not only drops yield but also affects grain quality which is shown in Fig1. It spreads across the field from plant to another plant through air [3], [4].

    KeywordsRice; Leaf disease detection; Machine Learning; Deep Learning; Artificial Intelligence


      Today across the world and specifically in India Rice is considered as best source of food. During the cultivation process the Paddy crop, usually it passes through various diseases in different stages of cultivation. Through early detection of such diseases and remedial steps taken timely can avoid huge loss and can yield good crop that is high in quantity and best in quality. The best expert guidance in agriculture is not accessible in remote areas as time is not

  2. Leaf Blast(LB)

    Fig. 1: Brown Spot(BS) Disease

    sufficient to reach such remote locations.

    The aim of research in agriculture is to improve the productivity and quality of the crop yielded with less expenditure and good yield. The crop can be managed effectively with the timely diagnosis of the disease and implementation of the possible solution within the best possible time for effective control over the factors effecting the productivity and quality of the crop. Diagnosing manually is a hectic task as it involves various parameters. Hence the implementation of Automated Systems is a mandate requirement to reach the farmers to help them in early detection of the diseases with improved accuracy. During this process the advanced techniques of machine learning play a key role in the disease classification process [1], [2].


      Oryza leaves affected with a disease can cause damage to plant to a great extent which can lessen the yield. These diseases are mainly caused because of bacteria, fungi and viruses with their infection spread rapidly once affected which in fact affects the entire crop if not timely diagnosed. The various diseases which affect rice crop are numerous.

      It is due to Magnaporthe oryzae, which affects the crop by forming lesions on the plant leaves and also on other areas of the plant like stems, roots, seeds, etc., [5], [6].

      Unlike the brown spot, the patches found on the leaves in this case are boat shaped, with centre in grey along with a thick outline of brown as shown in Fig. 2.

      Fig. 2: Rice Leaf Blast Disease

  3. Sheath Blight(SB)

Symptoms noticed in initial stage of sheath blight can be seen near water level on sheaths of leaf. Greenish grey spots on the sheath of leaf either in oval or elliptical are formed irregularly (Fig. 3). The enlarged spots become grey combined with white with an outline border in purple brown or blackish brown can be seen. The infection on the plants upper portion spread rapidly from waterline which reaches the flag end of the leaf. Similarly the infection penetrates into the inner sheath which leads to the death of the plant.

Majorly the said sheath blight can be suspected more in older plants and sheaths aged five to six weeks, this affects the crop at large where the grain is produced with weak formation. The major reason behind the above problem is due to excess consumption of fertilizers containing nitrogen [3].

E. Sheath Rot(SR)

The formation of small seized black lesions found on the sheath of outer leaf close to water line[3], which spread to the sheath of inner leaf resulting in rotting of tissues. It is shown in the below Fig. 5.

When the crop is infected by SR we can observe several spots on the leaf. Spots overtime enlarge, merge or grow along with the crop covering almost entire leaf sheath. Panicles may partly come out or retain inside the sheath. Leaflets that are affected may contain a good amount of white powdery fungal development called mycelium, appearing on the external surface[6].

Fig. 3: Sheath Blight Disease

D. Bacterial Leaf Blight(BLB)

The bacterial leaf blight is bacterium which penetrates through hyadathodes cutting wounds in the tip of leaf leading to the death of seedling (Fig. 4). The wounds enlarge with margin in wave shape which turns the straw into yellow within few days. The lesions turn the entire leaf into white or straw coloured upon the progress of the disease. The wounds may be seen on sheath of leaves, the dew drops with bacterial masses can be seen on fresh lesions early in the morning[1],[6].

Fig. 4: Bacterial Leaf Blight Disease

Fig. 5: Sheath Rot Disease

F. Leaf Smut(LS)

Leaf smut, it can be identified by the vertical scratches found on leaf blades in slight black colour, because of LS even the leaf tips may change to grey and become dry[7][8].

In fact it is not a considerable major disease but it can create a scope for other diseases by its nature of creating an environment that can promote and encourage the growth of fungi. The wounds of LS on the leaves may be in oval in shape or circular in shape or irregular shape with a kind of rough surface as depicted in below Fig. 6[9].

Fig. 6: Leaf Smut Disease


    Plant Diseases are recognized efficiently with the help of automated systems. The techniques used for processing ranges from application of various Image Processing, Computer Vision, Fuzzy Logic, Soft Computing, ML, DL and many more.

    Method presented by authors in [6] is used to diagnose and correctly classify given leaf samples into LB, BS, BLB and SR diseases. Significant features are extracted by processing the images and which are then used for classification with the help of MDC and k-NN resulting with accuracy of 0.89 and

    0.87 respectively.

    Authors of [10] provided with comparative analysis of various techniques such as segmentation, clustering and classifiers based on ANNs, Naives Bayes, Fuzzy Logic, PSO, Membership Function, Combined classifiers, OPDPA, Evaluation Theory on the basis of Minimum Path, Fractal Dimensions, SVM, Discriminant Analysis, AdaBoost Algorithm and Rule Set Theory for finding diseases of different plants with accuracy ranging from 0.70 to 1.0.

    Soluion given by [11], applies image processing and ANN mechanisms to detect diseases in various commonly grown plants resulting with an accuracy of 0.89. Technique of [12], detects diseases of Beans and Tea plants applying image processing followed by back propagation NN.

    Proposed technique of [13] has two phases. Phase-I deals with model training which includes leaf images gathering, preprocessing, feature extraction followed by ANN. Phase-II is testing phase which includes preprocessing steps followed by K-Means for segmentation process followed by classification with ANN. At last grading of disease is performed based on the amount of defected portion with the help of Fuzzy Logic.

    Paper presented in [14], provides a comparative study of various image processing along with soft computing techniques for disease recognition and classification into following four categories, Fungal, General, Bacterial, and Deficiency classes where accuracy ranges from 0.90 to 1.0.

    Different ML/ DL techniques were presented to detect diseases especially in rice plants by the papers [15] to [24], [25], [26], and [27] for which accuracy spans from 0.73 to 1.0.


Application of advanced Machine Learning methods has simplified the process of disease type recognition at the early stage accurately. In this section the most recent proposed solutions that are outperforming for types of disease identification of O. Sativa crops are presented along with their performance measure (Table I).

Diseases Identified



Performan ce Measure and score

Referred Paper



and Tungro

Deep feature based SVM

Accuracy: 0.97

F1 score: 0.98

Prabira Kumar Sethy et al(2020) [15]


LB, False smut, BS, Bakanae disease, SB, SR, BLB,

Bacterial sheath rot,Seeding blight and Bacterial wilt

Deep CNN

Accuracy: 0.95

Yang et al(2017) [16]




Mean Accuracy:

Long Tian et al(2021) [17]

Diseases Identified



Performan ce Measure and score

Referred Paper



and Tungro

Deep feature based SVM

Accuracy: 0.97

F1 score: 0.98

Prabira Kumar Sethy et al(2020) [15]


LB, False smut, BS, Bakanae disease, SB, SR, BLB,

Bacterial sheath rot,Seeding blight and Bacterial wilt

Deep CNN

Accuracy: 0.95

Yang et al(2017) [16]




Mean Accuracy:

Long Tian et al(2021) [17]

TABLE I: Comparative Analysis of most outstanding ML/ DL Algorithms for leaf disease identification of O. Sativa



BLB, LB, and SB

CNN model with high level fusion

Accuracy: 1.0

Lei Feng et al(2020) [18]


LB, Red blight, Stripe blight, and SB

CNN with SVM

Accuracy: 0.96

F. Jiang et al(2020) [19]


a)LS, BLB and BS

Extreme gradient boosting(XGBoo st)

Accuracy: 0.86

F1 score: 0.87

Muhammad Anwarul Azim et al(2020) [9]



Probabilistic Neural Network(PNN)

Accuracy: 0.91

F1 score: 0.92

Li-Wei Liu et al(2021)[20]


BS, BLB and LS

AlexNet neural network

Accuracy: 0.99

Md. Mafiul Hasan Matin et al(2020)[7]


LS, BLB and BS

Decision Tree(j48)

Accuracy: 0.97

F Score: 0.97

Kawcher Ahmed et al(2019) [8]


Stackburn, bacterial leaf streak, False smut, LS, Leaf scald, White tip, LB, Stem rot, Sheath spot, SR, Grain spotting and peck, Kernel smut, and SB

Deep_Transfer Learning

Accuracy: 0.98

Junde Chen et al(2020)[21]


LB, BLB and BS

CNN_Transfer Learning

Accuracy: 0.92

Shreya Ghosal et al(2020)[1]


LS and BS


Accuracy: 0.79

P A Gunawan et al (2021)[22]


LB, BS and Hispa


Mean Accuracy: 0.987

Bari et al(2021)[23]


SB, LB,False Smut, BS, and Stem Rot.

Minimum Distance Classifier(MDC)

Accuracy: 0.81

Vikas Sharma et al(2020)[3]



and SR

Deep Neural Network_ Jaya Optimization Algorithm(DNN


Mean Accuracy: 0.95

Ramesh et al(2019)[24]

It is observed that Deep Learning method CNN with its variations such as deep_CNN, CNN_SVM, CNN_High level Fusion, CNN_Transfer Learning, Faster_RCNN are performing better in terms of testing accuracy. Further it is noticed that the basic ML technique, the Decision Tree is outperforming with an accuracy of 0.97 detecting LS, BLB and BS diseases. The AlexNet_Neural Network has shown significant improvement in testing accuracy for diseases LS, BLB and BS.

Techniques of [2] and [10] are covering wide span of diseases while providing promising accuracy of 0.95 and 0.98 respectively. Fig. 7 shows F1-score along with Accuracy of the solutions presented in [1], [6], [7] and [9]. It is perceived that solution of [1] is expressing better F1-score than all other solutions and at same time maintaining the accuracy that of solution [9] which is the maximum observed accuracy among others.

Fig. 7: Comparative Analysis of ML/DL Techniques based on F1-Score

and Accuracy


Rice crop suffers from various types of diseases and their detection at early stage can be automated with a better accuracy using application of recent proposed Machine Learning and Deep Learning techniques. This paper presents more recent and outperforming solutions. It is observed that CNN model with high level fusion technique is the best solution with test accuracy exhibiting 1.0 for the three common diseases of rice plant i.e., LS, BLB and BS. The solution AlexNet_Neural Network resulting with 0.99 accuracy standing as second best technique over other contemporary outperforming solutions. It is also observed that, Deep feature based SVM method is exhibiting better F1- score along with Accuracy among other methods.


  1. S. Ghosal and K. Sarkar, "Rice Leaf Diseases Classification Using CNN With Transfer Learning," 2020 IEEE Calcutta Conference (CALCON), 2020, pp. 230-236, doi: 10.1109/CALCON49167.2020.9106423.

  2. Santanu Phadikar and Jaya Sil, Rice Disease Identification using Pattern Recognition Techniques, Proceedings of 11th International Conference on Computer and Information Technology (ICCIT 2008) 25-27 December, 2008. DOI:10.1109/ICCITECHN.2008.4803079

  3. Vikas Sharma, Aftab Ahmad Mir , Dr. Abid Sarwr, Detection of Rice Disease Using Bayes Classifier and Minimum Distance Classifier, Journal of Multimedia Information System VOL. 7, NO. 1, March 2020 (pp. 17-24): ISN 2383-7632 (Online)

  4. Hiroyuki Sato, Ikuo Ando, Hideyuki Hirabayashi, Yoshinobu Takeuchi, Sakae Arase, Junichi Kihara, Hiroshi Kato, Tokio Imbe, Hiroshi Nemoto, QTL analysis of brown spot resistance in rice (Oryza sativa L.), Breeding Science, 2008 Volume 58 Issue 1 Pages 93-96,

  5. Wen-Liang Chen, Yi-Bing Lin, Fung-Ling Ng, Chun-You Liu, Yun- Wei Lin, RiceTalk: Rice Blast Detection using Internet of Things and Artificial Intelligence Technologies, in IEEE Internet of Things Journal, vol. 7, no. 2, pp. 1001-1010, Feb. 2020, doi: 10.1109/JIOT.2019.2947624.

  6. Amrita A. Joshi, B.D. Jadhav, Monitoring and Controlling Rice Diseases Using Image Processing Techniques, 2016 International Conference on Computing, Analytics and Security Trends (CAST) College of Engineering Pune, India. Dec 19-21, 2016.

  7. Md. Mafiul Hasan Matin, Amina Khatun, Md. Golam Moazzam, Mohammad Shorif Uddin, An Efficient Disease Detection Technique of Rice Leaf Using AlexNet, Journal of Computer and

    Communications, Vol.8 No.12(2020), Paper ID 105804, 9 pages,


  8. Kawcher Ahmed, Tasmia Rahman Shahidi, Syed Md. Irfanul Alam and Sifat Momen, Rice Leaf Disease Detection Using Machine Learning Techniques, 2019 International Conference on Sustainable Technologies for Industry 4.0 (STI), 24-25 December, Dhaka

  9. Muhammad Anwarul Azim, Mohammad Khairul Islam, Md. Marufur Rahman, Farah Jahan, An effective feature extraction method for rice leaf disease classification, TELKOMNIKA Telecommunication, Computing, Electronics and Control Vol. 19, No. 2, April 2021, pp. 463~470 ISSN: 1693-6930, accredited First Grade by Kemenristekdikti, Decree No: 21/E/KPT/2018 DOI: 10.12928/TELKOMNIKA.v19i2.16488

  10. Gittaly Dhingra1, Vinay Kumar1, Hem Dutt Joshi, Study of digital image processing techniques for leaf disease detection and classification, Multimed Tools Appl Nov 2017,

  11. Diptesh Majumdar , Dipak Kumar Kole , Aruna Chakraborty , Dwijesh Dutta Majumder, REVIEW: DETECTION & DIAGNOSIS OF PLANT LEAF DISEASE USING INTEGRATED IMAGE PROCESSING APPROACH, International Journal of Computer Engineering and Applications, Volume VI, Issue-III June 2014

  12. Malti K. Singh , Subrat Chetia, Detection and Classification of Plant Leaf Diseases in Image Processing using MATLAB, International Journal of Life Sciences Research ISSN 2348-3148 (online) Vol. 5, Issue 4, pp: (120-124), Month: October – December 2017, Available at:

  13. A. Rastogi, R. Arora and S. Sharma, "Leaf disease detection and grading using computer vision technology & fuzzy logic," 2015 2nd International Conference on Signal Processing and Integrated Networks (SPIN), 2015, pp. 500-505, doi: 10.1109/SPIN.2015.7095350.

  14. Hiteshwari Sabrol, Satish Kumar, Recent Studies of Image and Soft Computing Techniques for Plant Disease Recognition and Classification, International Journal of Computer Applications (0975 8887) Volume 126 No.1, September 2015

  15. Prabira K Sethy, Nalini Kanta Barpanda, Amiya Kumar Rath, Santi Kumari Behera, Deep feature based rice leaf disease identification using support vector machine, Computers and Electronics in Agriculture,

    Volume 175, 2020,105527, ISSN 0168-1699,

  16. Yang Lu, Shujuan Yi, Nianyin Zeng, Yurong Liu, Yong Zhang, Identification of rice diseases using deep convolutional neural networks,

    Neurocomputing, Volume 267, 2017, Pages 378-384, ISSN 0925-2312,

  17. Long Tian, Bowen Xue, Ziyi Wang, Dong Li, Xia Yao, Qiang Cao, Yan Zhu, Weixing Cao, Tao Cheng, Spectroscopic detection of rice leaf blast infection from asymptomatic to mild stages with integrated machine learning and feature selection, Remote Sensing of Environment, Volume 257, 2021, 112350, ISSN 0034-4257, om/science/article/pii/S0034425721000687)

  18. Feng L, Wu B, Zhu S, Wang J, Su Z, Liu F, He Y and Zhang C (2020) Investigation on Data Fusion of Multisource Spectral Data for Rice Leaf Diseases Identification Using Machine Learning Methods. Front. Plant Sci. 11:577063. doi: 10.3389/fpls.2020.577063.

  19. Feng Jiang, Yang Lu, Yu Chen, Di Cai, Gongfa Li, Image recognition of four rice leaf diseases based on deep learning and support vector machine, Computers and Electronics in Agriculture,Volume 179, 2020, 105824, ISSN 0168-1699,

  20. Liu, L.-W.; Hsieh, S.-H.; Lin, S.-J.; Wang, Y.-M.; Lin, W.-S. Rice Blast (Magnaporthe oryzae) Occurrence Prediction and the Key Factor Sensitivity Analysis by Machine Learning. Agronomy 2021, 11, 771.

  21. Junde Chen, Defu Zhang, Yaser A Nanehkarana and Dele Li, Detection of rice plant diseases based on deep transfer learning, J Sci Food Agric 2020; 100: 32463256 ( DOI 10.1002/jsfa.10365

  22. P A Gunawan, E N Kencana and K Sari, Classification of rice leaf diseases using artificial neural network, Journal of Physics: Conference

    Series, Volume 1722, Tenth International Conference and Workshop on High Dimensional Data Analysis (ICW-HDDA-X), 2021 J. Phys.:

    Conf. Ser. 1722 012013

  23. Bari BS, Islam MN, Rashid M, Hasan MJ, Razman MAM, Musa RM, Ab Nasir AF, P.P. Abdul Majeed A. 2021. A real-time approach of diagnosing rice leaf disease using deep learning-based faster R-CNN framework. PeerJ Computer Science 7:e432

  24. S. Ramesh , D. Vydeki, Recognition and classification of paddy leaf diseases using Optimized Deep Neural network with Jaya algorithm, Information Processing in Agriculture, Volume 7, Issue 2, 2020, Pages 249-260, ISSN 2214-3173, ( 9)

  25. A. Sony, Prediction of Rice Diseases Using Convolutional Neural Network (in Rstudio), International Journal of Innovative Science and Research Technology, Volume 4, Issue 12, December 2019

  26. Basavaraj S. Anami, Naveen N. Malvade, Surendra Palaiah, Deep learning approach for recognition and classification of yield affecting paddy crop stresses using field images, Artificial Intelligence in Agriculture,Volume 4,2020,Pages 12-20,ISSN 2589- 7217,

  27. Lakshmana Rao Rowthu, M Naveen Kumar, Ch Narayana Rao, Early Identification of Rice Plant Diseases using Machine Learning Algorithms, Journal of Information and Computational Science, Volume 10 Issue 9 2020, pg 368 – 372

Leave a Reply