Deep Neural Networks Based Recognition of Plant Diseases by Leaf Image Classification.

The latest generation of convolutional neural networks (CNNs) has achieved impressive results in the field of image classification. This paper is concerned with a new approach to the development of tomato plant disease recognition model, based on leaf image classification, by the use of deep convolutionalnetworks. Novel way of training and the methodology used facilitate a quick and easy system implementation in practice. The developed model is able to recognize different types of tomato plant diseases out of healthy leaves, with the ability to distinguish plant leaves from their surroundings. According to our knowledge, this method for plant disease recognition has been proposed for the first time. All essential steps required for implementing this disease recognition model are fully described throughout the project, starting from gathering images in order to create a database, assessed by agricultural experts. Neural network, was used to perform the disease detection. The experimental results on the developed model achieved detection between 85% and 95%.

Strawberry plant is one of fruit plants that can grow in the highland with an altitude of 1000-1500m above sea level. Good care is taken to overcome diseases in strawberry plants. Farmers are usually difficult to distinguish the type of disease in strawberry plants. Deep learning is one of the ways to distinguish the types of diseases in plants by processing image feature extraction. Convolutional Neural Network (CNN) is one of the methods used in deep learning to classify diseases in strawberry plants by extracting image features of strawberry leaf disease. In this study, we propose a deep learning method using CNN to classification 6 types of diseases in strawberry plants with the use of 4663 strawberry leaf disease image data. The result of accuracy is 63,7%.

Objectives: The aim of the study is to automate the plant disease recognition and classification process by using image processing and soft computing techniques. Methods/Analysis: The proposed method examined the five types of tomato plant diseases using natural outdoor images in the study. The tomato plant images categorized into six categories including five disease infected that are bacterial leaf spot, fungal septoria leaf spot, bacterial canker, fungal lateblight, tomato leaf curl and one non-infected (healthy). The total 180 images of the dataset used for training and testing purpose. The total thirteen features computed by using CIE XYZ color space conversions that included color moments, histogram, and color coherence vector features. For classification, computed features are fed into three classifiers, i.e., “Fuzzy Inference System based on subtractive clustering”, “Adaptive neuro-fuzzy inference system using hybrid learning algorithm and multi-layer feed forward back propagation neural network” for classification of six injured and healthy tomato plant disease. Finding: The classification accuracy is best yielded with multi-layer feed forward back propagation classifier of 87.2%. Novelty/Improvement: Usually, in the studies the only one type of plant disease considered for the recognition and classification purpose. The current study considered five different types of tomato plant diseases including fungal, bacterial and viral. It indicates that the proposed algorithm could reliably classify the different types of plant diseases in digital images.

The accurate recognition and early warning for plant diseases and pests are a prerequisite of intelligent prevention and control for plant diseases and pests. As a result of the phenotype similarity of the hazarded plant after plant diseases and pests occur, as well as the interference of the external environment, traditional deep learning models often face the overfitting problem in phenotype recognition of plant diseases and pests, which leads to not only the slow convergence speed of the network, but also low recognition accuracy. Motivated by the above problems, the present study proposes a deep learning model EResNet-support vector machine (SVM) to alleviate the overfitting for the recognition and classification of plant diseases and pests. First, the feature extraction capability of the model is improved by increasing feature extraction layers in the convolutional neural network. Second, the order-reduced modules are embedded and a sparsely activated function is introduced to reduce model complexity and alleviate overfitting. Finally, a classifier fused by SVM and fully connected layers are introduced to transforms the original non-linear classification problem into a linear classification problem in high-dimensional space to further alleviate the overfitting and improve the recognition accuracy of plant diseases and pests. The ablation experiments further demonstrate that the fused structure can effectively alleviate the overfitting and improve the recognition accuracy. The experimental recognition results for typical plant diseases and pests show that the proposed EResNet-SVM model has 99.30% test accuracy for eight conditions (seven plant diseases and one normal), which is 5.90% higher than the original ResNet18. Compared with the classic AlexNet, GoogLeNet, Xception, SqueezeNet and DenseNet201 models, the accuracy of the EResNet-SVM model has improved by 5.10%, 7%, 8.10%, 6.20% and 1.90%, respectively. The testing accuracy of the EResNet-SVM model for 6 insect pests is 100%, which is 3.90% higher than that of the original ResNet18 model. This research provides not only useful references for alleviating the overfitting problem in deep learning, but also a theoretical and technical support for the intelligent detection and control of plant diseases and pests. © 2024 Society of Chemical Industry.

The latest generation of convolutional neural networks (CNNs) has achieved impressive results in the field of image classification. This paper is concerned with a new approach to developing a plant disease recognition model, based on leaf image classification, by the use of deep convolutional networks. A novel way of training and the methodology used to facilitate a quick and easy system implementation in practice. The developed model can recognize three types of diseases of one plant and pesticides and/or fertilizers are advised according to the severity of the diseases. The type of green leaf disease is recognized by CNN. After recognition, the predictive remedy is suggested that can help agriculture-related people and organizations to take appropriate actions against these diseases. Internet of Things (IoT) is a technology that allows things to communicate and connect with each other. Change the patterns and processes in both industry and agriculture towards higher efficiency. An intelligent farming system (IF) to improve the production process in planting. IF composes of two main parts which are a sensor system and a control system. The control part which watering and roofing systems of an outdoor farm based on the statistical data sensed from the sensor systems. A set of decision rules based on the sensed data is developed to automatically make a decision on whether the watering and roofing system should be on or off. As the energy demand and the environmental problems increase, natural energy sources have become very important as an alternative to conventional energy sources. The renewable energy sector is fast gaining ground as a new growth area for numerous countries.

Strawberry is a plant that has many benefits and a high risk of being attacked by pests and diseases. Diseases in strawberry plants can cause a decrease in the quality of fruit production and can even cause crop failure. Therefore, a method is needed to assist farmers in identifying the types of diseases in strawberry plants. Currently, there are many methods to assist farmers in identifying types of disease in plants, including strawberry plants. In this study, a system is proposed to be able to detect strawberry plant diseases by classifying the disease based on healthy and diseased strawberry leaf images. The proposed system is the Convolutional Neural Network (CNN) algorithm using MobileNetV3-Large and EfficientNet-B0 models to train pre-processed datasets. The results of this study obtained the best accuracy reaching 92.14% using the MobileNetV3-Large architecture with the hyperparameter optimizer RMSProp, epochs 70, and learning rate 0.0001. The percentage of the evaluation model using MobileNetV3-Large for precision, recall, and F1-Score achieved 92.81%, 92.14%, and 92.25%. Whereas in the EfficientNet-B0 architecture, the best accuracy results only reach 90.71% with the hyperparameter optimizer Adam, 70 epochs, and a learning rate of 0.003. Then, the precision, recall, and F1-scores for EfficientNet-B0 reached 92.65%, 90.00%, and 90.37%. Overall, it presents fairly good results in classifying strawberry leaf plant disease. Furthermore, in future work, it needs to obtain higher accuracy by generating more datasets, trying other augmentation techniques, and proposing a better model.

The existence of pests and diseases in plants and crops has a substantial impact on agricultural production within a country. Monitoring plants meticulously to detect and identify diseases is a common practice among farmers and experts. However, this approach is often laborious, costly, and not entirely reliable. To mitigate this issue, we propose a Disease Recognition Model based on leaf image classification. Our objective is to detect plant diseases using image processing techniques, specifically leveraging Convolutional Neural Networks (CNNs). CNNs are a category of artificial neural networks tailored for handling pixel-based inputs, particularly adept at image recognition tasks. Abbreviation- CNN (Convolutional Neural Network), SVM (Support Vector Machine), KNN (K Nearest Neighbour), ANN (Artificial Neural Network), GPU (Graphics processing unit)

The way a doctor can predict what kind of diseases a patient is suffering from, similarly, the fastest stratagem of predicting plant diseases is to analyze leaf's physiognomy changes and compare them with their actual color, shape, structure, etc. Plant disease recognition on the basis of leaf's physiognomy changes is the fundamental purpose of our project. We have used Convolutional Neural Network as a training method. CNN works via 3 dimensions of layers where neurons of every layer aren't fully connected to the next layer rather only a small portion is connected and the output will be decreased to a single dimension. For this, even with big datasets CNN works faster than any other networks. That's why we have used it for achieving a satisfying accuracy outcome. The program will exert plant images as input and detaching them to predict plant diseases. So it will help to identify and differentiate various types of plant diseases like aster yellows, bacterial wilt, scab, etc. quite easily & correctly.

Plant health is important in maintaining the sustainability of the foods crop. The key to prevent the loss of yield of plant crops is the identification of plant diseases. The process of monitor plant health manually is challenging as it required expert knowledge which is expensive and time-consuming. Hence, the image processing techniques can be useful for the detection and classification of plant leaf disease. In this project, the leaf images of 5 plant types in the PlantVillage dataset are used for plant type and plant disease classification. The original images are resized to the required input sized and the proposed background removal methods (improved HSV and GrabCut segmentation) are performed to reduce the background noise. The segmented images are then given to proposed models (AlexNet and DenseNet121) for training and classification. For plant type classification, DenseNet121 got a better validation accuracy of 99% compared to AlexNet with 91.2%. After that, the leaf image is given to plant disease models according to their species. All the plant disease models training with DenseNet121 can achieve high validation accuracy of 99%, 99%, 100%, 100% and 97% for apple, grape, potato, strawberry and tomato. Lastly, a user-friendly graphical user interface (GUI) is developed.

Recent advances in computer vision have led to the development of a robust learning technique that can identify and diagnose plant diseases using photos captured by a camera. This practical approach can help detect various illnesses in different plant species, including apples, corn, grapes, potatoes, tomatoes, and sugar cane. The system's architecture specifically targeted these plants for detection and recognition, and it can detect several plant diseases. To develop deep learning models for plant disease detection and recognition, scientists used 35,000 photos of both disease-free and diseased plant leaves. The system achieved up to 100% accuracy in identifying the type of plant and the diseases affecting it, with the trained model achieving an accuracy rate of 96.5%. The technique involved using convolutional neural networks, computer vision, deep learning, and plant disease recognition.

Abstract— Plants and crops that are infected by pests have an impact on the country's agricultural production. Usually, farmers or professionals keep a close eye on the plants in order to discover and identify diseases. However, this procedure is frequently time- consuming costly, andimprecise. Plant disease detection can be done by looking for a spot on the diseased plant's leaves. The goal of this paper is to create a Disease Recognition Model that is supported by leaf image classification. To detect plant diseases, we are utilizing image processing with a Convolution neural network (CNN). A convolutional neural network(CNN) is a form of artificial neural network that is specifically intended to process pixel input and is used in image recognition.Farmers do not expertise in leaf disease so they produce less production. Plant leaf diseases detection is the important because profit and loss are depends on production. Index Terms: Image processing,Crops,Support vector Machine,Plant disease,Classification. Keywords— Heart Disease, Diabetes, Machine Learning.

Automating the detection of different types of diseases in plants is one of the most complex recent challenges faced by agricultural experts all over the world. Cassava is loaded with carbohydrates and is mostly cultivated by small farmers in Sub-Saharan Africa as a security crop as it is capable of growing amidst drought which affects mostly all the other crops directly and adversely. At present, farmers are completely dependent on plant health experts to come and have a weekly or bi-weekly random inspection of the plants. They collect leaf samples from 3 different parts of the plant – top, middle, and lower parts and send them to the lab for testing. This takes up a lot of time and does not leave enough time for farmers to find a solution once the plants get affected. Therefore, it is important for farmers to detect the diseases as early as possible. To tackle this problem, we propose a disease detection model for cassava plants using a Convolutional Neural Network (CNN). The features in the cassava plant images which signify the presence of disease will be automatically learned by the Neural Network. The farmers cannot detect these features manually. This research was conducted using a dataset of 21,397 labeled images collected during a regular survey in Uganda comprising four diseases namely Cassava Bacterial Blight (CBB), Cassava Brown Streak Disease (CBSD), Cassava Green Mottle (CGM), Cassava Mosaic Disease (CMD) and the dataset also contains healthy images of the plant. Most of the images in this dataset were taken by farmers in their gardens to provide a realistic representation of how farmers would diagnose these diseases in real life. These images were annotated by experts at the National Crops Resources Research Institute (NaCRRI) in collaboration with the AI lab at Makerere University, Kampala.

1. Tianjin University of Science and Technology, Tianjin, China 2. Tianjin Agricultural University, Tianjin, China,300384 3. Mokpo National University, Muan, Republic of Korea 4. Jeonbuk National University, Jeonju, Republic of Korea 5. Lushan Botanical Garden, Chinese Academy of Sciences,Jiangxi Province,China Plant diseases and pests cause significant losses to farmers and threaten food security worldwide. Monitoring the growing conditions of crops and detecting plant diseases is critical for sustainable agriculture. Traditionally, crop inspection has been carried out by people with expert knowledge in the field. However, regarding any activity carried out by humans, this activity is prone to errors, leading to possible incorrect decisions. Innovation is, therefore, an essential fact of modern agriculture. In this context, deep learning has played a key role in solving complicated applications with increasing accuracy over time, and recent interest in this type of technology has prompted its potential application to address complex problems in agriculture, such as plant disease and pest recognition. Although substantial progress has been made in the area, several challenges remain, especially those that limit systems to operate in real-world scenarios. This research topic aims to explore recent advanced AI methods for plant disease and pest recognition for real-world applications.In this topic editing, a total of 21 papers have been published, encompassing the contributions of 84 distinct authors. The content primarily delves into the realms of disease detection and identification pertaining to crops such as tomatoes, peppers, rice, corn, soybeans, alongside fruits including apples, grapes, and blueberries. Furthermore, the scope encompasses the automation of harvesting processes and anomaly detection within economically significant crops like tea and tobacco. Notably, there are also inquiries into the behavioral patterns of Diptera Tephritidae.The focus of these studies centers on the identification, detection, and segmentation of plant diseases, as well as aspects related to harvesting, pests, and growth monitoring. Among the research papers, 18 are dedicated to the study of plant diseases, while the remaining cover various other topics. The primary subjects of investigation include leaves, fruits, and flowers of plants, with additional examinations into pests and fungi. Specifically, there are 17 papers primarily focused on leaf analysis, 3 on fruit analysis, and 1 each on pests, flowers, and fungi. From this, it is evident that the prevailing research methodologies predominantly focus on the recognition of plant diseases through leaf analysis.From a task-oriented perspective, the requirements for identification tend to be idealized, often limited to utilizing images containing a single leaf. However, as research and applications progress, achieving segmentation, detection, and optimized deployment in real-world scenarios becomes increasingly crucial.Consequently, there is a burgeoning interest in exploring the vast research potential surrounding fruits, flowers, pests, and fungi.In identification tasks, the primary challenge lies in the feature extraction capacity of models, mainly due to the visual resemblance of different diseases In detection tasks, the primary challenges continue to arise from dataset limitations. This is because the task closely simulates real farm environments, leading to issues such as unknown classes, annotation errors, and labeling inaccuracies within the dataset. Additionally, the challenge of detecting small yet densely packed objects persist. Dong et al. proposed two approaches to address these challenges. Firstly, they introduce a method that utilizes teacher-student networks for self-supervised repair of imprecise and incomplete annotations.Secondly, they present an Open-World detection method (Dong, et al.) In conclusion, regardless of identification, segmentation, or detection tasks, dataset limitations remain the primary bottleneck in the development of advanced AI methods for plant disease and pest recognition. As articulated in the perspective article (Xu et al.), imperfect datasets always entail additional risks and challenges. However, effectively leveraging these datasets can still reduce costs and enhance efficiency. For research topics in plant disease and pest recognition that rely heavily on domain expertise, further refining the objectives of recognition and expanding the applicability of models to minimize data dependency in realworld environments are two key objectives for future development.

Throughout the area of image processing, the new generation of convolutional neural networks CNNs hasproduced remarkable performances. The paper reflects on a new approach to the production of utilizingdeep convolutionary networks of a model of the identification of plants centered on the picture classificationof the surface. Throughout reality, a modern model of teaching and technique allows it simple and fast tointroduce the program. With the ability to differentiate the plant leaves from the environment, the engineeredmodel will recognise 13 specific forms of plant diseases from safe leaves. This approach for identifyingplant disease was introduced for the first time, according to our understanding. The entire paper outlinesall important measures possible for the introduction of this model for disease identification, beginning withthe picture collection to establish a database reviewed by agricultural experts. The comprehensive CNNpreparation was carried out by Caffe, a fundamental research system developed by Berkley Vision andLearning Center. On average, the experimental results of the built model were 91 to 98 percent reliable forseparate class research.

Every year India losses the significant amount of annual crop yield due to unidentified plant diseases. The traditional method of disease detection is manual examination by either farmers or experts, which may be time-consuming and inaccurate. It is proving infeasible for many small and medium-sized farms around the world. To mitigate this issue, computer aided disease recognition model is proposed. It uses leaf image classification with the help of deep convolutional networks. In this paper, VGG16 and Resnet34 CNN was proposed to detect the plant disease. It has three processing steps namely feature extraction, downsizing image and classification. In CNN, the convolutional layer extracts the feature from plant image. The pooling layer downsizing the image. The disease classification was done in dense layer. The proposed model can recognize 38 differing types of plant diseases out of 14 different plants with the power to differentiate plant leaves from their surroundings. The performance of VGG16 and Resnet34 was compared. The accuracy, sensitivity and specificity was taken as performance Metrix. It helps to give personalized recommendations to the farmers based on soil features, temperature and humidity