A Review: CNN Based Plant Disease Detection

Abstract

Identifying plant diseases is the initial step toward mitigating losses in agricultural product yield and quantity. This work focuses on exploiting such technical breakthroughs, with an actual focus on plant disease detection. The system will analyze plant photos using cutting-edge image processing algorithms to precisely diagnose diseases, gauge the extent of damage, and make wise suggestions for necessary pesticides and nutrient supplements. The suggested remedy will identify particular nutrient deficits causing problems with crop health in addition to detecting the existence of illnesses. The system will aid farmers in making well-informed decisions about how to use nutrients and pesticides by integrating a thorough analysis. This will ultimately help to manage crop diseases in a timely and efficient manner, maximize agricultural yield, and reduce financial losses.

Introduction

I.INTRODUCTION

India primarily relies on agriculture, with approximately 70% of its population dependent on this sector. Crop loss is a major concern for the country and countries economy and productivity. Leaves are delicate part of the plant so the disease symptoms first shown on leaves [21]. Early disease identification is essential because of variety of crops and the requirement for appropriate insecticides. Farmers used to rely solely on labour intensive, human observation, which meant that professional monitoring was needed. Yet, effective substitutes have surfaced in the form of improved computerized and partially computerized determination of plant diseases in past several years.

It has been demonstrated that disease detection by symptom analysis on plant leaves is more precise, cost-effective, and quicker than manual observation [28]. This strategy is especially important because disease symptoms frequently show up on fruit, leaves, and stems. In order to give farmers a useful tool for early and accurate monitoring, especially in situations where they might not have a thorough understanding of crop diseases. For solving this scenario, numerous researchers from all over the globe have developed cutting-edge systems to facilitate autonomous identification of plant diseases using different kinds of machine learning (ML) [22, 26, 24, 25] and deep learning (DL) approaches [23, 27]. So, here we discussed the research on identifying plant diseases that has been proposed by several researchers.

 This work focuses on exploiting such technical breakthroughs, with a actual focus on plant disease detection. This technology attempts to give farmers a comprehensive tool that not only recognizes diseases but also provides customized recommendations for required pesticides and nutrients by utilizing the power of sophisticated algorithms. This proactive approach aims to stipulate farmers with timely information, enabling them to make informed decisions and implement precision agricultural techniques, thus improving crop resilience and overall agricultural output.

II. BACKGROUND CONCEPT

A. Machine Learning (ML)

ML is an artificial intelligence (AI) advance in technology which offers systems the capacity to instinctively study from experience and get improve at it without needing to be obviously designed. The formation  of software application which retrieve data further utilize to learn on their own is important target of ML. Learning begin with data or inspection, like examples, command to discover trends or modification in data and use the examples we provide to guide future decisions.

The central objective is to enable computers to study by itself , without lending a hand  or without the participation of human being. Plant disease classification is aided by ML. Utilizing this technique is seen as an important first move toward eliminating plant diseases, and it has also raised crop production [29]. Experts are able to determine the source of plant illnesses and analyze them with the support of a variety of ML and DL techniques [4]. Among the machine learning (ML) techniques utilized in identifying illnesses decision-making are support vector machines (SVM) and K-means [31].

1. SVM Classifier

SVM that is Support Vector Machine (SVM) is actually an supervised ML approach. It is can be applied for regression and classification. It can be used to solve the binary and multi-class classification problems depicted in Figs. (2) and (3) [32]. Two varieties of SVM are nonlinear and linear SVM. Different classes are created from the provided image by using linear or straight-line SVM. When linear lines cannot be used to classify the images, nonlinear SVM is used [33]. The main focus  is actually to recognize hyperplane in the N-dimensional space which is used to split the data points into different feature space groups.

Thus, SVM generates a set of hyperplanes [35] or a single hyperplane [34]. These hyperplanes are utilize to categorize the data points into different group  [36]. determining the best hyperplane to split the classes maximally and to perform well on data that has not been seen before [37].

2. KNN Classifier

The number of nearest neighbors that are taken into consideration while making predictions is represented by the letter "K" in KNN [38]. In machine learning, K-Nearest Neighbors has been employed in classification, statistical estimation, and pattern recognition [39]. The algorithm makes use of a distance metric, such as the Euclidean distance, in determining how similar two data points are [38] [40].

3. Deep Learning (DL)

DL is actually within area of ML [41], often used for the detection of objects [42, 43,51], picture classification, language processing [44, 45,52], and processing of natural languages [46, 47]. In the current the context, real-time applications and deep learning architectures for disease identification have become a focus [46]. Deep learning involves use of deep neural networks, which consist of multi-layered neural networks. It has capacity to figure out intricate links and patterns in data.

4. CNN

A specific category of ML known as CNN (Convolutional Neural Network) is a DL mechanism. To examine graphical information CNN extremely use. CNN constructed with the capability to automatically and appropriately identify features of the spatial hierarchy from the input data [48]. It consist of numerous layers. Out of every layer each layer conducts a specific role during the feature extraction and classifying process [47].

5. Convolution Layer

Most of computations originates in the convolutional layer of CNN, which is the principle structural component. For finding particular feature in this layer filter that is a relatively small weight matrix which is float across the receptive part of the input image.

6. Pooling Layer

Pooling layer is most important section right after the convolution layer. Like the convolution layer, it carried out task of cleaning of input images. But it additionally offers distinct role. Pooling layer lowering overall dimension of arriving data yet preserving the crucial data in order to enhance the whole accuracy of network.

7. Fully Connected Layer  

This layer categorizes images depending upon feature acquired from prior layers. So that’s why this layer is very important in latter phases of  CNN. When a neuron in individual layer is said to be fully linked, it implies that every other neuron in the layer below it is connected as well. Various properties extracted from the former layers have been combined and entrusted to particular groups or outcomes from fully connected layer. In this layer all activation unit comprises an attachment for each input from prior layers, it enable the CNN to test whole features and then sorting the data.

III.LITERATURE REVIEW

In this section, various methods for the identification of disease in plants are reviewed. Researchers have done a lot of studies on plant disease detection. Here are some works done by different authors.

1. DL-based methods have been proposed for realistic detection & identification of insect in soybean crops, as reported by Turkey D, Singh KK, et al. (2023). To evaluate the reliability of this method in insect detection accuracy, the performance of various transfer learning (TL) techniques was examined. The suggested approached achieved an efficacy of 98.75% with YoloV5, 97% with InceptionV3, & 97% with CNN. Among these, the YoloV5 algorithm proved to be efficient, operating at 53 frames per second in dynamic detection. Moreover, a data of agricultural diseases was compiled & annotated using a variety of imaging technologies. This approach simplified the study, reduced the workload for producers, & yielded superior results.
2. Ahmed I. & Yadav PK. (2023) have illustrated two viral, two mold diseases four bacterial infections & one mite-related illness using the "PlantVillage" dataset. Pictures of unaffected leaves from a twelve crop variety were also included. Machine learning approached like SVM, CNN, & GLCM were utilized for developing prediction models. The progression of AI in classification has paralleled the development of Artificial Neural Networks (ANNs) through back propagation. In extension of investigation  Real-time leaf photos were used for illness identification, employing K-nearest neighbors (KNN) in addition to machine learning techniques. It observed that overall efficacy of 96%, 94%, 95%, & 97% for tomato plants, 99% & 98% for apple & rice plants, respectively. Precision, recall, & F-measure metrics assessed multi-layer classification, utilizing a dataset with one symptom pool per class.
3. Algani YMA, Caro OJM, et al.(2022)  introduced an adaptive deep learning approach for disease identification & bifurcation termed Ant Colony Optimization with Convolutional Neural Network (ACO-CNN). ACO was utilized to evaluate disease diagnosis in plant leaves, while CNN classifiers were employed to eliminate texture, geometric plant location, color, texture from the provided photos. Various effectiveness metrics demonstrate that this approach outperforms previous methods in accuracy. These metrics are utilized for analysis & to propose a recommended approach. Concrete measures were taken to implement these strategies. Ultimately, in the form of accuracy, precision, recall, & F1-score, the ACO-CNN model outperformed C-GAN, CNN, & SGD models. The ACO-CNN model achieved 99.98% accuracy.
4. Dai G, Fan J, et al. (2023) introduced PPLCNet, a deep learning model incorporating GAP layers, a multi-level attention mechanism, & dilated convolution. Innovative meteorological data augmentation was utilized to exp& the sample size, enhancing feature extraction resilience & generalization. The model addresses insufficient data information extraction by extending the perceptual field of convolutional domains through saw-tooth dilated convolution with configurable expansion rates. A lightweight CBAM attention mechanism in the middle layer enhances information representation. Overfitting is mitigated by the GAP layer, reducing parameter quantity & complexity. During validation, the PPLC-Net model achieved a recognition accuracy of 99.702% & an F1-score of 98.442% on the retained test dataset. With 15.486 million parameters & 5.338 billion FLOPs, the model satisfies requirements for accurate & rapid recognition.
5. P. Nayar & S. Chhibber (2022) explored a novel method by employing leaf classification through deep convolutional networks (DCN) for creating disease detection model. The exp&ing field of computer presents facility to enhance & broaden the utilization of precision crop protection techniques, opening new avenues applications in precision agriculture. This methodology enables swift & direct implementation of the system. The study utilized a data containing 77,000 photos of both healthy & diseased plant leaves. By training a CNN to classify diseases in plant & discern their presence, another model was trained to detect diseases using YOLOv7. The advanced model achieved precision & recall rates of 65%, 59%, & 65%, while the trained classification model attained an efficacy of 99.5%.
6. A CNN solution based on DL was given by Kukadiya H. & Meva D. in 2022 for classification & differentiation of cotton leaf diseases. Various studies have been conducted on common leaf diseases in various crops; however, this researched provided a reliable & effective method to diagnosing leaf diseases in cotton. Three cotton leaf diseases if not discovered were successfully classified & identified by the suggested approach. CNN is the suggested model, with training & testing efficacy of 100% & 90%, respectively for identification & classification of cotton leaf diseases.
7.  Adesh V. Panchal, Subhash Ch&ra Patel, et al. (2023) propose a process for identifying crop diseases by labeling affected leaves according to disease patterns, followed by pixel-based techniques to enhance information extraction from processed images. Feature extraction, image segmentation, & crop disease classification are then conducted based on identified patterns implementing a convolutional neural network (CNN). A public dataset comprising around 87,000 RGB-type photos, including healthy & diseased leaves, is utilized for demonstration.
8.  Sami Ur, Fakhre Alam, Rahman, et al. (2023) introduce an image processing-based techniques for automated diagnosis & preventive measures of leaf diseases in tomato crops. Their approach utilizes the Gray Level Co-Occurrence Matrix (GLCM)  approach of    tomato leaves 13 statistical variables , which  classified into various diseases using Support Vector Machines (SVM). Experimental results demonstrate high efficacy rates: 100% for healthy leaves, 95% for early blight, 90% for septoria leaf spots, & 85% for late blight. This method is operationalized as a mobile application.
9. Gangwar A, Rani G, & Dhaka VPS. (2023) propose an image segmentation-based, DL technique for tomato disease detection. Their approach utilizes the VIA tool for leaf masks, a adaptive U-Net model for image segmentation, & a convolution network for disease classification. With a 98.12% accuracy rate, this method shows promise for automatic tomato disease detection, potentially enhancing yield & reducing crop loss.
10. B. Paulos & M. M. Woldeyohannis (2022) give an innovative approach for identifying & categorized disease of coffee plant, stating that it is essential for increased output. Here, 1120 photos from the Wolaita Sodo Agricultural Research Center were utilized to train the DL model. To address data overfitting, an augmentation technique was employed. 3360 photos in all were utilized. To achieve the best results when classifying these diseases, they have contrasted training from scratch with TL techniques. This leads to a 98.5% accuracy rate for training from scratch & 97.01 & 99.89 efficacy rates for transfer-based learning using Mobilnet & Resnet50, respectively. When it came to classifying photos, the pre-trained Resnet50 model best than other alternative techniques.
11. Shoaib M. et al. (2022) propose a DL-based method for diseases identification of tomato plant, utilizing the InceptionNet model & a CNN architecture. With more than 18,000 segmented & non-segmented tomato leaf images, they trained the CNN using a supervised learning approach.by implementing U-Net & Modified U-Net segmentation algorithms; they achieved superior performance with the modified U-Net model, reached 98.66% IoU score & 98.73% on the dice.
12. Attallah O. (2023) presents a algorithm for autonomous diseases identification of tomato plant, by using three compact CNNs. By extracting deep features signals from the CNNs' last linked layer & incorporating components from all three architectures, they achieved high accuracy rates with KNN & SVM classifiers, reaching 99.92% & 99.90% accuracy, respectively.
13.  Ksibi A. & Ayadi M. (2022) introduce a deep feature concatenation (DFC) technique, utilizing ResNet50 & MobileNet models to create the MobiResNet neural network. With a dataset of 5400 olive leaf photos captured by an agricultural UAV, MobiResNet outperformed ResNet50 & MobileNet, achieving an overall classification accuracy of 97.08%.
14.  Albattah W., Nawaz M., et al. (2022) propose a robust method for classification of  plant diseases using a Custom CenterNet architecture with DenseNet-77. By incorporating improved key point extraction & utilizing the Plant Village Kaggle database, their method proves more reliable & efficient than recent techniques for plant disease recognition.
15.  Dharmendra Saraswat, Ahmad, Aanis, et al. (2023) conduct a comprehensive analysis of seventy research works on deep learning applications in agricultural disease diagnosis & control. Covering various aspects from dataset requirements to deep learning techniques, their review aims to guide future tool development & support farmers in managing plant diseases effectively.
16. Al-Gaashani MSAM, Shang F, et al. (year) propose a method for diseases identification of tomato plant using transfer learning (TL) & feature signals concatenation. Features are further extracted from pre-trained MobileNetV2 & NASNetMobile kernels & concatenated by implementing kernel principal component analysis. These concatenated features are then fed into a traditional learning system, where multinomial logistic regression (LR) achieves the highest accuracy of 97% among tested classifiers.
17. Amritha Haridasan, et al. (2023) introduces a computer vision-based methodology for identification of disease in rice plant, integrating DL, ML, & image processing tool techniques. This approach significantly reduces the reliance on traditional methods for safeguarding paddy crops against common diseases. By employing convolutional neural networks CNN & a support vector machine SVM  classifier, the proposed approach achieves a validation accuracy of 0.9145, aiding in timely diagnosis & treatment recommendations.
18. Poornima Singh Thakur (2023) presents the "VGG-ICNN," a lightweight CNN for identification of crop diseases by plant-leaf images.  Over using 6 million parameters, VGG-ICNN outperforms other high-performing deep learning models while covering a wide range of crop types across multiple datasets. Experimental results demonstrate high accuracy, particularly achieving 99.16% accuracy on the PlantVillage dataset & consistent performance across various crop datasets.
19.  Shoaib et al. (2022) highlight the effectiveness of CNN models in obtaining improved detection accuracy using imaging data for classification of  plant disease. CNNs excel in feature extraction & classification, outperforming other ML & DL techniques with accuracy rates ranging from 99% to 99.2%.
20. Upadhyay SK, Kumar A. (2022) propose an innovative method for analysing  & classifying rice plant diseases depend on lesion size, shape, & color in leaf images. Utilizing a fully CNN trained of 4000 image samples, the proposed method achieves a high efficacy rate of 99.7%, surpassing existing methods for plant disease identification & categorization.

Conclusion

It is clear from the thorough literature assessment that by using Image processing adaptive techniques on plant disease detection, scientists have made great progress toward creating precise & dependable methods for recognizing & categorizing diseases impacting different crops. Of the various ways investigated, a few st& out for their creative strategies, solid findings, & useful applications. From the above reviews, Algani YMA, Caro OJM, et al. (2022) introduced a noteworthy methodology name as: Ant Colony Optimization - Convolutional Neural Network (ACO-CNN) for detecting & classifying plant leaf diseases. By leveraging the strengths of both convolutional neural networks & ant colony optimization, their method achieves impressive accuracy rates by extracting crucial information from images. The ACO-CNN model surpasses in terms of accuracy, precision, recall, & F1-score not only traditional methods but also modern DL models like C-GAN, CNN, & SGD. Their approach demonstrates outst&ing performance & reliability in disease identification & classification, achieving an impressive accuracy 99.98% with mentioned evaluation metrics. The outcomes attained by Algani YMA, Caro OJM, et al. (2022) highlight how well their suggested strategy works to provide solution for the problems related to detection of plant disease. They have developed a strong framework which precisely identify & categorize illnesses by utilizing the complimentary characteristics of CNN & ACO. This framework provides important insights for crop management & precision agriculture methods. Furthermore, the application of sophisticated DL methods such as ACO-CNN represents a viable avenue for further research projects aiming at improving the effectiveness & dependability of plant disease detection systems. To sum up, the research conducted by Algani YMA, Caro OJM, et al. (2022) is noteworthy for its significant addition for detection of plant disease in agriculture sector. By combining conventional optimization methods with cutting-edge deep learning techniques to tackle intricate agricultural problems is demonstrated by their creative methodology & outst&ing performance metrics. Their work therefore forms the basis for future developments in the creation of dependable & effective methods for the detect ion & provide treatment of plant diseases in agriculture sector.

References

[1] Turkey D, Singh KK, Tripathi S. Performance analysis of AI-based solutions for crop disease identification detection, & classification. Smart Agric Technol. 2023. https://doi.org/10.1016/j.atech.2023.100238. [2] Ahmed I, Yadav PK. A systematic analysis of machine learning & deep learning based approaches for identifying & diagnosing plant diseases. Sustain Oper Comput. 2023;4:96–104. https://doi.org/10.1016/j.susoc.2023.03.001 [3] Algani YMA, Caro OJM, Bravo LMR, Kaur C, Al Ansari MS, Bala BK. Leaf disease identification & classification using optimised deep learning. Meas Sensors. 2023;25:100643. https://doi.org/10.1016/j.measen.2022.10064. [4] Dai G, Fan J, Tian Z, Wang C. PPLC-Net : neural network-based plant disease identification model supported by weather data augmentation & multi-level attention mechanism. J King Saud Univ Comput Inf Sci. 2023;35(5):101555.https://doi.org/10.1016/j.jksuci.2023.101555. [5] P. Nayar, S. Chhibber, & A. K. Dubey, “An Efficient Algorithm for Plant Disease Detection Using Deep Convolutional Networks,” Proceedings - 2022 14th IEEE International Conference on Computational Intelligence & Communication Networks, CICN 2022. pp. 156–160, 2022. [6] Kukadiya H, Meva D. Automatic cotton leaf disease classification & detection by convolutional neural network. Berlin: Springer Nature Switzerl&; 2022. [7] Panchal, Adesh V., Subhash Ch&ra Patel, K. Bagyalakshmi, Pankaj Kumar, Ihtiram Raza Khan, & Mukesh Soni. \"Image-based plant diseases detection using deep learning.\" Materials Today: Proceedings 80 (2023): 3500-3506. [8] Rahman, Sami Ur, Fakhre Alam, Niaz Ahmad, & Shakil Arshad. \"Image processing based system for the detection, identification & treatment of tomato leaf diseases.\" Multimedia Tools & Applications 82, no. 6 (2023): 9431-9445. [9] Gangwar A, Rani G, Dhaka VPS. detecting tomato crop diseases with Ai: leaf segmentation & analysis. Int Conf Trends Electron Inform. 2023. https://doi.org/10.1109/ICOEI56765.2023.10125891. [10] E. B. Paulos & M. M. Woldeyohannis. Detection & classification of coffee leaf disease using deep learning,” no. January 2023, 2022. https://doi.org/10.1109/ICT4DA56482.2022.9971300. [11] Shoaib M, et al. Deep learning-based segmentation & classification of leaf images for detection of tomato plant disease. Front Plant Sci. 2022;13:1–18. https://doi.org/10.3389/fpls.2022.1031748. [12] Attallah O. Tomato leaf disease classification via compact convolutional neural networks with transfer learning & feature selection. Horticulturae. 2023. https://doi.org/10.3390/horticulturae9020149. [13] Ksibi A, Ayadi M, Soufene BO, Jamjoom MM, Ullah Z. Movies-net: a hybrid deep learning model for detecting & classifying olive leaf diseases. Appl Sci. 2022. https://doi.org/10.3390/app122010278. [14] Albattah W, Nawaz M, Javed A, Masood M, Albahli S. A novel deep learning method for detection & classification of plant diseases. Complex Intell Syst. 2022;8(1):507–24. https://doi.org/10.1007/s40747-021-00536-1. [15] Ahmad, Aanis, Dharmendra Saraswat, & Aly El Gamal. \"A survey on using deep learning techniques for plant disease diagnosis & recommendations for development of appropriate tools.\" Smart Agricultural Technology 3 (2023): 100083. [16] Al-gaashani MSAM, Shang F, Muthanna MSA, Khayyat M, El-Latif AAA. Tomato leaf disease classification by exploiting transfer learning & feature concatenation”. IET Image Process. 2022. https://doi.org/10.1049/ipr2.12397. 65. [17] Haridasan, Amritha, Jeena Thomas, & Ebin Deni Raj. \"Deep learning system for paddy plant disease detection & classification.\" Environmental Monitoring & Assessment 195, no. 1 (2023): 120. [18] Thakur, Poornima Singh, Tanuja Sheorey, & Aparajita Ojha. \"VGG-ICNN: A Lightweight CNN model for crop disease identification.\" Multimedia Tools & Applications 82, no. 1 (2023): 497-520. [19] Shoaib, M., Hussain, T., Shah, B., Ullah, I., Shah, S. M., Ali, F., et al. (2022a). Deep learning-based segmentation & classification of leaf images for detection of tomato plant disease. Front. Plant Sci. 13. doi: 10.3389/fpls.2022.1031748. [20] Upadhyay SK, Kumar A. A novel approach for rice plant diseases classification with deep convolutional neural network. Int J Inf Technol. 2022;14(1):185–99. https://doi.org/10.1007/s41870-021-00817-5. [21] Zhou, R., Kaneko, S., Tanaka, F., Kayamori, M., Shimizu, M., ‘Disease detection of Cercospora Leaf Spot in sugar beet by robust template matching’, Computers and Electronics in Agriculture, Volume 108, pp. 58-70, 2014. [22] Ahmed, K., Shahidi, T.R., Irfanul Alam, S.M., Momen, S., 2019. Rice leaf disease detection using machine learning techniques. 2019 International Conference on Sustainable Technologies for Industry 4.0 (STI). IEEE, Dhaka, Bangladesh, pp. 1–5. [23] Chohan, M., Khan, A., Katper, S., Mahar, M., 2020. Plant disease detection using deep learning. Int. J. Recent Technol. Eng. 9 (1), 909–914. https://doi.org/10.35940/ijrte. A2139.059120. [24] Panda, Panigrahi Kshyanaprava, Himansu, Das, Kumar, Sahoo Abhaya, Chandra, Moharana Suresh, 2020. Maize leaf disease detection and classification using machine learning algorithms. Progress in Computing. Analytics and Networking. Springer Singapore, Singapore, pp. 659–669. [25] Mohameth, F., Bingcai, C., Sada, K.A., 2020. Plant disease detection with deep learning and feature extraction using Plant Village. J. Comp. Commun. 8 (6), 10–22. https://doi.org/ 10.4236/jcc.2020.86002. [26] Naik, M.R., Sivappagari, C.M.R., 2016. Plant leaf and disease detection by using HSV features and SVM classifier. Int. J. Eng. Sci. 6 (12), 1–4. [27] Rao, D.R., Krishna, M., Ramakrishna, B., 2020. Smart ailment identification system for Paddy crop using machine learning. Int. J. Innov. Eng. Manag. Res. 9 (3), 96–100. [28] Barbedo, J.G.A., ‘A review on the main challenges in automatic plant disease identification based on visible range images’, Biosystems Engineering, Volume 144, pp. 52-60, 2016 [29] Rehan Ullah Khan , Khalil Khan, Waleed Albattah, and Ali Mustafa Qamar,“Learning to Deep Learning Journey ” Hindawi, Wireless Communications and Mobile Computing ,Volume 2021, Article ID 5541859, https://doi.org/10.1155/2021/5541859 [30] R. I. Hasan, S. M. Yusuf, and L. Alzubaidi, “Review of the state of the art of deep learning for plant diseases: a broad analysis and discussion,” Plants, vol. 9, no. 10, article 1302, 2020. [31] S. Wallelign, M. Polceanu, and C. Buche, “Soybean plant disease identification using convolutional neural network,” in Proc. Thirty-First International Florida Artificial Intelligence Research Society Conference (FLAIRS-31), pp. 146–151, Melbourne, FL, USA, 2018. [32] Mathur, A., & Foody, G. M. (2008). Multiclass and binary SVM classification: Implications for training and classification users. IEEE Geoscience and Remote Sensing Letters, 5(2), 241–245. https://doi.org/10.1109/LGRS.2008.915597 [33] J. A. Wani et al., “Machine Learning and Deep Learning Based Computational Techniques in Automatic Agricultural Diseases Detection: Methodologies, Application and Challenges,” Archives of Computational Methods in Engineering, vol. 29, no. 1, pp. 641-677, 2022. https://doi.org/10.1007/s11831-021-09588-5 [34] Institute of Electrical and Electronics Engineers. (n.d.). 2019 3rd International Conference on Imaging, Signal Processing and Communication (ICISPC) : July 27-29, 2019, Singapore. [35] Ahmed, I., & Yadav, P. K. (2023). A systematic analysis of machine learning and deep learning based approaches for identifying and diagnosing plant diseases. Sustainable Operations and Computers, 4, 96–104. https://doi.org/10.1016/j.susoc.2023.03.001 [36] Nikith, B. V., Keerthan, N. K. S., Praneeth, M. S., & Amrita, D. T. (2022). Leaf Disease Detection and Classification. Procedia Computer Science, 218, 291–300. https://doi.org/10.1016/j.procs.2023.01.011 [37] Iniyan, S., Jebakumar, R., Mangalraj, P., Mohit, M., & Nanda, A. (2020). Plant Disease Identification and Detection Using Support Vector Machines and Artificial Neural Networks. Advances in Intelligent Systems and Computing, 1056, 15–27. https://doi.org/10.1007/978-981-15-0199-9_2 [38] Barman, U., & Choudhury, R. D. (2022). Smartphone assist deep neural network to detect the citrus diseases in Agri-informatics. Global Transitions Proceedings, 3(2), 392–398. https://doi.org/10.1016/j.gltp.2021.10.004 [39] Shruthi Su, Nagaveni Veerakyatharayappa, B.K. Raghavendra (2019), “ A Review on Machine Learning Classification Techniques for Plant Disease Detection ”, Conference: 2019 5th International Conference on Advanced Computing & Communication Systems (ICACCS), DOI:10.1109/ICACCS.2019.8728415. [40] Zhang, B., Song, C., Li, Y., & Jiang, X. (2022). Spatiotemporal prediction of O3 concentration based on the KNN-Prophet-LSTM model. Heliyon, 8(11). https://doi.org/10.1016/j.heliyon.2022.e11670. [41] Li Deng and Dong Yu (2014), “ Deep Learning : Methods and Applications\", Foundations and Trends® in Signal Processing : Vol.7 : No. 3-4, pp 197-387. https://dx.doi.org/10.1561/2000000039. [42] Mads Dyrmann, Henrik Kastoft, Henrik Skov Midtiby (2016), “Plant Species Classification using deep convolutional neural network” Biosystems Engineering, volume 151, https://doi.org/10.1016/j.biosystemseng.2016.08.024. [43] Geetharamani, G.; Pandian, A. Identification of plant leaf diseases using a nine-layer deep convolutional neural network. Comput. Electr. Eng. 2019, 76, 323–338. [44] Huang, G.; Liu, Z.; Laurens, V.; Weinberger, K.Q. Densely connected convolutional networks. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, Honolulu, HI, USA, 21–26 July 2017; pp. 4700–4708. Available Online : https://openaccess.thecvf.com/content_cvpr_2017/papers/Huang_Densely_Connected_Convolutional_CVPR_2017_paper.pdf (accessed on 22 July 2021). [45] He, K.; Zhang, X.; Ren, S.; Jian, S. Identity Mappings in Deep Residual Networks. In European Conference on Computer Vision; Springer: Cham, Switzerland, 2016. [46] Sk Mahmudul Hassan, Arnab Kumar Maji, “Plant Disease Identification Using a Novel Convolutional Neural Network”, IEEE ACCESS, Digital Object Identifier 10.1109/ACCESS.2022.3141371. [47] Singh, S. P., Pritamdas, K., Devi, K. J., & Devi, S. D. (2022). Custom Convolutional Neural Network for Detection and Classification of Rice Plant Diseases. Procedia Computer Science, 218, 2026–2040. https://doi.org/10.1016/j.procs.2023.01.179 [48] Subramani, M. (n.d.). Plant Leaf Disease Detection Using Convolution Neutral Network. https://www.researchgate.net/publication/372676472 [49] Raheel Muzzammel , Ali Raza (2020), “A Support Vector Machine Learning-Based Protection Technique for MT-HVDC Systems”, research gate, DOI:10.3390/en13246668. [50] Phung, V.H.; Rhee, E.J. A Deep Learning Approach for Classification of Cloud Image Patches on Small Datasets. J. Inf. Commun. Converg. Eng. 2018, 16, 173–178, doi:10.6109/jicce.2018.16.3.173. [51] Ritika Wasan(2018) , “Deep Learning : Evolution and expansion ”,Cognitive Systems Research, volume 52, https://doi.org/10.1016/j.cogsys.2018.08.023. [52] Edna Chebet Too , Li Yujian , Sam Njuki , Liu Yingchun (2019), “A comparative study of fine-tuning deep learning models for plant disease identification”,Computers and Electronics in Agriculture, volume 161, https://doi.org/10.1016/j.compag.2018.03.032.

Copyright

Copyright © 2024 Rasika D. Shelke, Dr. Dinesh V. Rojatkar. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.