International Research Journal of Engineering and Technology (IRJET) Volume: 10 Issue: 07 | Jul 2023
www.irjet.net
e-ISSN: 2395-0056 p-ISSN: 2395-0072
Plant Leaf Disease Detection and Classification Using Image Processing Anmol Sinha1, Kartik Kishore2, Rahul Kumar Gupta3, Rajat Maheshwari4, Saksham Vats5, Parampreet Kaur6 1Student, Dept. of Computer Science and Engineering, Chandigarh University, Punjab, India 2Student, Dept. of Computer Science and Engineering, Chandigarh University, Punjab, India
3Student, Dept. of Computer Science and Engineering, Chandigarh University, Punjab, India 4Student, Dept. of Computer Science and Engineering, Chandigarh University, Punjab, India 5Student, Dept. of Computer Science and Engineering, Chandigarh University, Punjab, India
6Assistant Professor, Dept. of Computer Science and Engineering, Chandigarh University, Punjab, India
-------------------------------------------------------------------------***-----------------------------------------------------------------------research paper opens with an analysis of the significance of Abstract - The identification and classification of plant leaf
classifying and detecting plant diseases, then provides an overview of conventional techniques and discusses their drawbacks. Following that, the paper discusses image processing methods and how they might be used to recognise and categorise plant diseases. There is also discussion of the several steps in the image processing process, such as image capture, pre-processing, feature extraction, and classification.
diseases using image processing is a crucial field of study that has attracted a lot of interest lately. Large-scale losses to the agricultural business can be avoided with the help of early diagnosis and correct classification of plant diseases. In this study, we give a thorough analysis of the state-of-the-art methods for image-based diagnosis and categorization of plant leaf diseases. We go over numerous methods for feature extraction, image processing, and machine learning that are employed in this context. Additionally, we compare various approaches and talk about their benefits and drawbacks. We finish by outlining possible future directions for this area of study.
With accuracy levels ranging from 80% to 99.8%, depending on the approach taken, several studies have found promising results in this area. While the use of deep learning techniques, such as convolutional neural networks, has shown great promise and has accuracy levels of up to 99.8%, texture-based feature extraction and machine learning algorithms have been widely used and have been shown to be accurate up to 98.8% in some studies.
Key Words: Plant leaf disease, image processing, disease detection, plant disease classification, machine learning.
1. INTRODUCTION
However, a number of variables, such as the dataset chosen, the type of plant disease being diagnosed, and the image quality, determine how accurate these algorithms are. The effectiveness of the model can also be influenced by the machine learning algorithm and feature extraction technique employed.
Agriculture sector is seriously threatened by diseases of the plant leaves which drastically reduce crop yield and quality. For prompt and efficient disease treatment, it is essential to identify plant leaf diseases early and classify them correctly. Traditional techniques of illness identification rely on subjective and time-consuming visual evaluation by human experts. Additionally, it is not always easy for human experts to correctly differentiate between various plant diseases. Therefore, there is a need for automated systems that can reliably and effectively identify and classify plant leaf diseases.
Agriculture has a lot of opportunity to become more productive and sustainable by detecting and classifying plant leaf diseases using image processing. The creation of creative solutions to the problems the agriculture business is facing can be aided by further research in this field.
2. RELATED STUDIES
Images of plant leaves can be analysed, and illnesses can be found using image processing. In order to categorise the photos into various disease groups, image processing method can be used to draw features from the images. Models can be trained on these features, and new photos can be classified using machine learning algorithms as well.
For the detection and categorization of plant leaf diseases using image processing, numerous methods have been developed. Utilising colour-based features for disease detection is one method. RGB, HSV, and YCbCr are a few examples of colour spaces that can be used to extract colour-based information. A colour histogram method was utilised by Wang et al. [1] to identify illness in grapevine leaves. Their accuracy was 91.11% using a dataset of 90 photos.
With the aid of image processing, we present a summary of the numerous methods and algorithms used to identify and classify plant leaf diseases in this work. The
© 2023, IRJET
|
Impact Factor value: 8.226
|
ISO 9001:2008 Certified Journal
|
Page 34