Informatics Publishing Limited

Neelima Priyanka Nutulapati ¹, G. Adiline Macriga ², Dileep Pulugu ³, R. Thiru Murugan ⁴

Affiliations
1 Department of Information Technology, SRK Institute of Technology, IN
2 Department of Information Technology, Sri Sairam Engineering College, IN
3 Department of Computer Science and Engineering, Malla Reddy College of Engineering and Technology, IN
4 Department of Business Administration, Kalasalingam Academy of Research and Education, IN

Abstract

Plant diseases are a major threat to agriculture, particularly impacting the yield and quality of tomato crops. Early detection and accurate classification of these diseases are essential for effective management and mitigation. Traditional methods of disease detection are often labor-intensive and time-consuming. Although individual convolutional neural networks (CNNs) have shown promise in automated plant disease detection, their accuracy and robustness can be limited when used in isolation. This study proposes an ensemble learning approach that combines three state-of-the-art CNN architectures: AlexNet, ResNet50, and VGG16. A comprehensive dataset of tomato leaf images, categorized into bacterial, viral, fungal diseases, and healthy leaves, was used. Images were preprocessed and augmented to improve model generalization. Each model was trained separately, and their outputs were integrated using a weighted averaging mechanism to form the ensemble model. The weights for each model were optimized based on validation performance. The ensemble model significantly improved classification accuracy compared to individual models. The combined approach achieved an overall accuracy of 97.5%, with precision, recall, and F1-score exceeding 95% for all disease categories. Specifically, the accuracy for detecting bacterial diseases was 96.8%, viral diseases 97.2%, and fungal diseases 97.9%. The ensemble method demonstrated superior robustness and reliability in classifying diverse disease symptoms.

Keywords

Plant Disease Detection, Ensemble Learning, AlexNet, ResNet50, VGG16

Full Text

   

Dileep Pulugu ¹, P. Anusha ², Ratan Rajan Srivastava ³, R. Kalaivani ⁴, Subharun Pal ⁵

Affiliations
1 Department of Computer Science and Engineering, Malla Reddy College of Engineering and Technology, IN
2 Department of Electronics and Communication Engineering, R.M.K. Engineering College, IN
3 Department of Computer Science and Engineering, B.N. College of Engineering and Technology, IN
4 Department of Computer Science, National College, IN
5 SSM Research Center, Swiss School of Management, CH

Abstract

Video surveillance systems play a crucial role in ensuring public safety and security. However, the traditional methods of surveillance often fall short in effectively identifying individuals, particularly in crowded or dynamic environments. This research addresses the limitations of conventional video surveillance by proposing a machine learning-based facial recognition system. The increasing demand for robust security measures necessitates the development of advanced technologies in video surveillance. Facial recognition has emerged as a promising solution, but existing systems struggle with accuracy and efficiency. This research aims to bridge these gaps by leveraging machine learning techniques for facial recognition in video surveillance. Conventional video surveillance struggles with accurate and rapid identification of individuals, leading to potential security lapses. This research addresses the challenge of enhancing facial recognition accuracy in real-time video feeds, especially in scenarios with varying lighting conditions and occlusions. While facial recognition has gained traction, there is a significant research gap in the implementation of machine learning algorithms tailored for video surveillance. This study aims to fill this void by proposing a novel methodology that combines deep learning and computer vision techniques for robust facial recognition in dynamic environments. The proposed methodology involves training a deep neural network on a diverse dataset of facial images to enable the model to learn intricate facial features. Additionally, computer vision algorithms will be employed to handle challenges such as occlusions and varying lighting conditions. The model's performance will be evaluated using real-world video surveillance data. Preliminary results demonstrate a significant improvement in facial recognition accuracy compared to traditional methods. The machine learning-based system exhibits enhanced performance in challenging scenarios, showcasing its potential for practical implementation in video surveillance systems.

Keywords

Facial Recognition, Machine Learning, Video Surveillance, Deep Learning, Computer Vision

Full Text