Leaf Disease Detection System

Abstract

Agricultural productivity is something on which economy highly depends. This is the one of the reasons that disease detection in plants plays an important role in agriculture field, as having disease in plants are quite natural. If proper care is not taken in this area then it causes serious effects on plants and due to which respective product quality, quantity or productivity is affected. For instance a disease named little leaf disease is a hazardous disease found in pine trees in United States. Detection of plant disease through some automatic technique is beneficial as it reduces a large work of monitoring in big farms of crops, and at very early stage itself it detects the symptoms of diseases i.e. when they appear on plant leaves. This paper introduces an efficient approach to identify healthy and diseased or an infected leaf using image processing and machine learning techniques. Various diseases damage the chlorophyll of leaves and affect with brown or black marks on the leaf area. These can be detected using image prepossessing, image segmentation. Support Vector Machine (SVM) is one of the machine learning algorithms is used for classification. The Convolutional Neural Network (CNN) resulted in a improved accuracy of recognition compared to the SVM approach.

Introduction

I. INTRODUCTION

In India, for economic development, agriculture is a valuable source. To increase the production of food, the agriculture industries keep on searching for efficient methods to protect crops from damages. This makes researchers search for new efficient, and precise technologies for high productivity. The diseases on crops give low production and economic losses to farmers and agricultural industries. For a successful farming system, one of the essential things is disease identification. In general, by using eye observations, a farmer observes symptoms of disease in plants that need continuous monitoring. Different types of disease kill leaves in a plant. For identifying these diseases, farmers get more difficulties. For disease detection, the image processing methods are suitable and efficient with the help of plant leaf images. Though continuously monitoring of health and disease detection of plant increase the quality and quantity of the yield, it is costly. Machine learning algorithms are experimented due to their better accuracy. However, selection of classification algorithms appears to be a difficult task as the accuracy varies for different input data. The objectives are to detect leaf disease portion from the image, extract features of an exposed part of the leaf, and recognize diseased leaf through SVM. Further, Convolutional Neural Network is evaluated and compared for accuracy. The paper is arranged into five sections: the first section gives the introduction, the second section presents the literature survey, the third section discusses methodologies like feature extractions of images, SVM and CNN, the fourth section shows the result of classification, and the fifth section is about the conclusion and future scope.

What is Image Processing:-

Image processing has been proved to be effective tool for analysis of images in various fields and applications. Agriculture Sector where the parameters like canopy, yield, quality of product were the important measures from the farmer’s point of view. Many times the availability of expert and their services may consume a lot of time as well as expert advice may not be affordable. Image processing along with availability of communication network can change the situation of getting the expert advice well within time and at affordable cost since image processing was the effective tool for analysis of parameters. This paper intends to focus on the survey of application of image processing in agriculture field such as imaging techniques, leaf disease detection. The analysis of the parameters has proved to be accurate and less time consuming as compared to traditional methods. Application of image processing can improve decision making for vegetation measurement, irrigation, leaf sorting, etc. Irrigation/Water stress occurs when the water supply to the plants was limited Fertilizers, pesticides and quality of yield were the major factors of concern in agriculture. Most of the time the expertise were required to analyze the problems and which may be time consuming and costlier issue in developing countries. Image processing was one of the tools which can be applied to measure the parameters (leaf area index (LAI), nitrogen (N) uptake, total chlorophyll (Chl) content) related to agronomy with accuracy and economy. Applications of image processing in agriculture can be broadly classified in two categories: first one Remote Sensing depends upon the imaging techniques and second one based on applications like leaf disease Detection.

A. Project Idea

The agronomic requirements though in radically different ways to those currently used this has given rise to many new chances to service. So they should be tested via non-destructive techniques Leaves are delicate part of plant, The evaluation of agricultural harvest Classification is dynamic. The most important visual property is leafs texture and color. Hence, classification of leaf disease is necessary in evaluating agricultural produce, increasing market value and meeting quality standards. Identifying and taking further dealings for further diffusion of the diseases it is also helpful. The process will be too slow, If the identification and categorization is done through physical techniques, we need the experts help sometimes it will be error prone and who are less available. The labour’s classify based on color, size etc. if these quality methods are recorded into automatic system by using appropriate program design language then the effort will be error free and faster.

II. LITERATURE SURVEY

A. Survey Of Existing System

The existing method for plant disease detection is simply naked eye observation by experts through which identification and detection of plant diseases is done. For doing so, a large team of experts as well as continuous monitoring of plant is required, which costs very high when we do with large farms. At the same time, in some countries, farmers do not have proper facilities or even idea that they can contact to experts. Due to which consulting experts even cost high as well as time consuming too. In such conditions, the suggested technique proves to be beneficial in monitoring large fields of crops. Automatic detection of the diseases by just seeing the symptoms on the plant leaves makes it easier as well as cheaper. This also supports machine vision to provide image based automatic process control, inspection, and robot guidance

Plant disease identification by visual way is more laborious task and at the same time, less accurate and can be done only in limited areas. Whereas if automatic detection technique is used it will take less efforts, less time and become more accurate. In plants, some general diseases seen are brown and yellow spots, early and late scorch, and others are fungal, viral and bacterial diseases. Image processing is used for measuring affected area of disease and to determine the difference in the color of the affected area

Image segmentation is the process of separating or grouping an image into different parts. There are currently many different ways of performing image segmentation, ranging from the simple thresholding method to advanced color image segmentation methods. These parts normally correspond to something that humans can easily separate and view as individual objects. Computers have no means of intelligently recognizing objects, and so many different methods have been developed in order to segment images. The segmentation process is based on various features found in the image. This might be color information, boundaries or segment of an image [11], [13]. We use Genetic algorithm for color image segmentation.

C. ???????Problem Statement And Objective

Agricultural productivity is something on which economy highly depends. This is the one of the reasons that disease detection in plants plays an important role in agriculture field, as having disease in plants are quite naturalCrop diseases are a major threat to food security, but their rapid identification remains difficult in many parts of the world due to the lack of the necessary infrastructure Owing to changing climatic conditions, crops often get affected, as a result of which agricultural yield decreases drastically. If the condition gets worse, crops may get vulnerable towards infections caused by fungal, bacterial, virus, etc. diseases causing agents. The method that can be adopted to prevent plant loss can be carried out by real-time identification of plant diseases.To detect unhealthy region of plant leaves. Classification of plant leaf diseases using various features. Coding is used to analyze the leaf infection. The main aim is to obtain an automate detection of the plant diseases. On one hand visual analysis is least expensive and simple method, it is not as efficient and reliable as others .So, we detect the plant diseases using image processing and deep learning with greater efficiency

???????D. Project Scope

Plant diseases cause a major production and economic losses in the agricultural industry. The disease management is a challenging task. Usually the diseases or its symptoms such as colored spots or streaks are seen on the leaves of a plant. In plants most of the leaf diseases are caused by fungi, bacteria, and viruses.

The diseases caused due to these organisms are characterized by different visual symptoms that could be observed in the leaves or stem of a plant. Usually, these symptoms are detected manually. With the help of image processing, Automatic detection of various diseases can be detected with the help of image processing. Image processing plays a crucial role in the detection of plant diseases since it provides best results and reduces the human efforts.The scope of this project is very huge. It can be applied to various farming applications, and medicinal technologies. Farmers can use our app to detect the diseases present in their crops beforehand and can take preventive measures against it. Also in the field of medical science, it can be used to detect the diseases and get a knowledge of the percent of leaves that are affected by specific type of disease. By getting this analysis, we are able to prevent against those type of diseases which can happen in future.The image processing could be used in the field of agriculture for several applications. It includes detection of diseased leaf, stem or fruit, to measure the affected area by disease, to determine the color of the affected area. Tomato cultivation is one of the most remunerative farming enterprises in India.The naked eye observation by the experts is approach usually taken in identification and detection of plants. This approach is time consuming in huge farms or land areas. The use of image processing techniques in detection and identification of tomato plant diseases in the earlier stages and thereby the quality of the product could be increased

III. PROPOSED SYSTEM

Image acquisition is the first step of a plant disease detection system. By using digital cameras, scanners, or drones, high quality plant images can be captured. The images can also be taken from the web. Large numbers of image samples were collected from Kaggle datasets, which consists of diseased and healthy leaves. Image Pre-processing is used to increase the quality of leaf image and eliminate the unwanted noise.

The segmentation process is used to partition the plant image in various segments to separate the diseased portion of the leaf.

???????A. Algorithm

Following are the algorithms that are used in the proposed work. The Grey wolf optimization are used to optimize the features which are given by CNN.

1. CNN: Neural Network and Deep Learning How to find effective features is the core issue in image classification and pattern recognition. Humans have an amazing skill in extracting meaningful features, and a lot of research projects have been undertaken to build an FE system as smart as human in the last several decades. Deep learning is a newly developed approach aiming for artificial intelligence. Deep learning-based methods build a network with several layers, typically deeper than three layers. Deep neural network (DNN) can represent complicated data. However, it is very difficult to train the network. Due to the lack of a proper training algorithm, it was difficult to harness this powerful model until Hinton and Salakhutdinov proposed a deep learning idea [27]. Deep learning involves a class of models that try to learn multiple levels of data representation, which helps to take advantage of input data such as image, speech, and text. Deep learning model is usually initialized via unsupervised learning and followed by fine-tuning in a supervised manner.

The high level features can be learnt from the low-level features. This kind of learning leads to the extraction of abstract and invariant features, which is beneficial for a wide variety of tasks such as classification and target detection. There are a few deep learning models in the literature, including DBN [14], [15] SAE and CNN. Recently, CNNs have been found to be a good alternative to other deep learning models in classification and detection. In this paper, we investigate the application of deep CNN for HSI FE. The architecture of CNN is different from other deep learning models. There are two special aspects in the architecture of CNN, i.e., local connections and shared weights. CNN exploits the local correlation using local connectivity between the neurons of near layers. We illustrate this in Fig.4, where the neurons in the mth layer.

???????B. Data Model And Description: CNN Model

CNN is a widely-used image recognition model that has been shown to attain greater than 78.1% accuracy on the ImageNet dataset. The model is the conclusion of many ideas developed by many researchers over the years. An 256x256x3 input representing a visual field of 256 pixels and 3 color (RGB) channels. Five convolution layers, with a few interspersed max-pooling operations. Successive stacks of “CNN Models”. A softmax output layer at the end at an intermediate output layer just after the mixed layer. Steps involved in CNN are Convolution layer in CNN is performed on an input image using a filter. Relu (Rectified Linear Unit) which simply converts all of the negative values to 0 and keeps the positive values the same. Pooling layer is used to reduce the spatial size of the Convolved Feature. They are of two types such as Max Pooling and Average Pooling. Fully Connected layers in a neural networks is a layer where all the inputs from one layer are connected to every activation unit of the next layer. These networks are commonly trained under a log loss (or cross-entropy) system, giving a nonlinear variant of multinomial logistic regression.

???????C. Fundamental Model

The phases of the proposed work are:

1. Collection of raw data.
2. Training data using CNN model.
3. Testing data.
4. Determine accuracy.

???????D. Methodology

Image processing is a process of performing some operations on an image, in order to get an enhanced image or to extract some useful information from it. Acquisition, Segmentation, Augmentation, Feature Extraction. Image acquisition is the first main step of digital image processing. Image acquisition is simple process when given an image that is already in digital form. Generally, the image acquisition stage includes pre-processing, such as scaling etc. Image Segmentation procedures partition an image into its constituent parts or objects. Generally, autonomous segmentation is the most difficult tasks in digital image processing. A rugged segmentation process that brings a long way toward successful solution of imaging problems that require objects to be identified individually. Image Augmentation is a technique that is used to artificially expand the dataset, parameters that are generally used to increase the data sample count are zoom, shear, rotation, pre-processing function and so on. Image feature extraction is one of the most important segments in this project. Feature selection works by selecting the best features based on unilabiate statistical tests. The features are carefully selected based on their unique differences between the different types of leaves.

1. Support Vector Machine (SVM) Classifier: To classify the various diseases in plants any of the machine learning techniques can be used. The classification phase suggested deciding if the input image is healthy or diseased. In this paper Support Vector Machine (SVM) classifier has been used because it has some advantages over other classifiers such as effective in high dimensional spaces also in cases where the number of dimensions is greater than the number of samples. It is memory efficient as it uses a subset of training points in the decision function (called support vectors). SVM is a supervised machine learning algorithm used for both classification and regression. SVM is a discriminative classifier. In this approach, for classification, the SVM technique has been applied. In SVM, each data item is plotted as a point in n-dimensional space; the number of dimensions corresponding to the number of features being classified. The classification is obtained by discovering the hyper-plane that uniquely distinguishes between different groups of scattered data points. By finding the hyper-plane the classification is performed. Hyper-plane differentiates two classes very well.

2. Convolutional Neural Network (CNN): CNN is a class of deep neural networks. The CNN model comprises an input layer, convolution layer, pooling layer, a fully connected layer, and an output layer shown in figure 3. To classify the disease in plants in a precise manner the images are provided as input. The convolution layer is used for extracting the features from the images. The pooling layer computes the feature values from the extracted features.

Depending on the complexity of images, the convolution and pooling layer can be further increased to obtain more details. A fully connected layer uses the output of previous layers and transforms them into a single vector that can be used as an input for the next layer. The output layer finally classifies the plant disease.

???????

Conclusion

The existing method for plant disease detection is simply naked eye observation by experts through which identification and detection of plant diseases is done. For doing so, a large team of experts as well as continuous monitoring of plant is required, which costs very high when we do with large farms. In such conditions, the suggested technique proves to be beneficial in monitoring large fields of crops. Automatic detection of the diseases by just seeing the symptoms on the plant leaves makes it easier as well as cheaper and different diseases classification techniques used for plant leaf disease detection and an algorithm for image segmentation technique that can be used for automatic detection as well as classification of plant leaf diseases later. With very less computational efforts the optimum results were obtained, which also shows the efficiency of proposed algorithm in recognition and classification of the leaf diseases. This study summarizes major image processing used for identification of leaf diseases are k-means clustering, SVM. This approach can significantly support an accurate detection of leaf disease. By computing amount of disease present in the leaf, we can use sufficient amount of pesticides to effectively control the pests in turn the crop yield will be increased. We can extend this approach by using different algorithms for segmentation, classification. By using this concept the disease identification is done for all kinds of leaves and also the user can know the affected area of leaf in percentage by identifying the disease properly the user can rectify the problem very easy and with less cost.

References

[1] Sujatha R* , Y Sravan Kumar and Garine Uma Akhil, “leaf Disease Detection using Image Processing”,School of Information Technology and Engineering, VIT University, Vellore-Journal of Chemical and Pharmaceutical Sciences , 27th November 2017 [2] Thete Vaishali V , Thakare Pradnya R , Kadlag Gaurav B , [3] Prof. P.A. Chaudhari, “Leaf Disease Detection Using Image Processing”-BE Student, Electronics and Telecommunication Department, SVIT Chincholi, Nashik Professor, Information Technology Department, SVIT Chincholi, Nashik- February 2017 [4] Vijai Singh and A.K. Misra, “Detection of plant leaf diseases using image segmentation and soft computing techniques”-journal homepage: www.elsevier.com/locate/inpa-Computer Science Department, IMS Engineering College, Ghaziabad, UP, India and Computer Science & Engg. Department, MNNIT Allahabad, UP, India [5] Utkarsha N. Fulari , Rajveer K. Shastri , Anuj N. Fulari, “Leaf Disease Detection Using Machine Learning” -Journal of Seybold Report-Department of Electronics and Telecommunication, Vidya Pratishthan’s Kamalnayan Bajaj Institute of Engineering and Technology, Baramati, India [6] Anjana , Mr. Keshav Kishore , “Plant Leaf disease classification and detection with CNN”-2 Department of Computer Science & Engineering, AP Goyal Shimla University, INDIA - October - December 2018 [7] Keerthana M1, Raksha K R2, Thanuja V3, Sridhar R4 , “Classification and Identification of Leaf Diseases using Deep Learning”- 1,2,3Students, Department of Information Science & Engineering, Global Academy of Technology, Bengaluru, Karnataka, India [8] Asst. Professor, Department of Information Science & Engineering, Global Academy of Technology, Bengaluru, Karnataka, India-International Research Journal of Engineering and Technology (IRJET) - June 2020 [9] Emma Harte, “Plant Disease Detection using CNN”-School of Computing National College of Ireland Mayor Street, IFSC, Dublin 1 Dublin, Ireland- 7th September 2020. [10] Monishanker Halder, Ananya Sarkar, Habibullah Bahar, “PLANT DISEASE DETECTION BY IMAGE PROCESSING: A LITERATURE REVIEW”- Journal of food science technology- Department of Computer Science and Engineering, Jashore University of Science and Technology, Jashore-7408

Copyright

Copyright © 2023 Sahil Gupta, Vivek Pandey, Pravesh Pandey, Mukul Verma, Hasib Shaikh. 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.