International Research Journal of Engineering and Technology (IRJET)
e-ISSN: 2395-0056
Volume: 12 Issue: 05 | May 2025
p-ISSN: 2395-0072
www.irjet.net
IMAGE BASED PLANT DISEASE DETECTION A COMPARISON OF DEEP LEARNING AND CLASSICAL MACHINE LEARNING ALGORITHM Jenitha.R [1], Manonmani.D [2] , Priyanka.P [3] , Swetha.k [4] , Mr. K.Gopal M.Tech., [4] Student [1], Dept. of Information Technology, The kavery Engineering College Mecheri, Salem Professor [2], Dept. of Computer Science Engineering, The kavery Engineering College Mecheri, Salem
---------------------------------------------------------------------***--------------------------------------------------------------------effective, they often struggle with complex disease patterns, variations in lighting conditions, and diverse plant species. efficiency, driving to financial misfortunes and Deep learning, particularly Convolutional Neural Networks nourishment security concerns. Early and exact location (CNNs), has revolutionized image-based classification tasks of plant maladies is significant for viable administration by automatically learning hierarchical features from raw and avoidance. This paper presents a comparative images. CNNs have been successfully applied in various consider of classical machine learning and profound domains, including medical imaging, facial recognition, and learning calculations for plant malady location utilizing autonomous driving. In agriculture, CNNs eliminate the need image-based methods. The ponder investigates highlight for manual feature extraction by learning patterns directly extraction, classification precision, computational from plant images, making them more efficient and scalable effectiveness, and strength of different models. Classical for disease detection. Compared to traditional ML machine learning approaches, such as Bolster Vector approaches, CNNs can handle large-scale datasets and Machines (SVM)and Convolutional Neural Systems (CNNs) achieve higher accuracy in identifying plant diseases under . The test investigation illustrates that profound learning diverse environmental conditions. In this paper, we present models accomplish prevalent precision in plant malady a comparative study of deep learning-based CNNs and location whereas classical strategies offer interpretability classical ML techniques for image-based plant disease and computational productivity. The comes about detection. We analyze the effectiveness of different highlight the trade-offs between exactness and approaches in terms of classification accuracy, complexity, giving experiences into the best-suited computational efficiency, and practical applicability in smart approach for distinctive agrarian applications. agricultural systems. The goal is to provide insights into the strengths and limitations of each method, ultimately guiding Key Words: classical machine learning, classification, the selection of the most suitable approach for real-world precision and profound learning calculations deployment. I.INTRODUCTION
Abstract – Plant infections altogether affect rural
II. EXISTING SYSTEM
Agriculture is a vital sector that supports global food production and economic stability. However, crop diseases pose a significant challenge, leading to reduced yields, financial losses, and food insecurity. Early and accurate detection of plant diseases is crucial for effective disease management and prevention. Traditional methods of disease detection primarily involve manual inspection by farmers or agricultural experts, which is often time-consuming, laborintensive, and prone to human error. With the advancements in artificial intelligence (AI), machine learning (ML), and deep learning (DL), automated image-based plant disease detection has gained significant attention in recent years.Machine learning techniques have been widely used for plant disease classification by extracting handcrafted features from images and training models to differentiate between healthy and diseased plants. Classical ML algorithms such as Support Vector Machines (SVM), Decision Trees (DT), k-Nearest Neighbors (k-NN), and Random Forest (RF) have shown promising results in detecting plant diseases. These algorithms rely on predefined features such as texture, shape, and color, which require careful selection and manual tuning. While ML-based approaches have been
© 2025, IRJET
|
Impact Factor value: 8.315
Early plant disease detection systems used image processing techniques such as color segmentation, edge detection, and texture analysis to extract relevant features. These features included color histograms, shape descriptors, and gradientbased features that were then used for classification. Although these techniques provided a basic level of disease identification, they often lacked robustness in real-world conditions due to variations in lighting, background noise, and image quality.
2.1 Disadvantages
|
Feature Dependence: The accuracy of classical methods is constrained by the effectiveness of manual feature extraction.
Scalability Issues: These approaches struggle when applied to large, diverse datasets with multiple disease categories.
ISO 9001:2008 Certified Journal
|
Page 952