ICTACT Journal on Image and Video Processing

D. C. Jullie Josephine ¹, B. Yuvaraj ², Sathesh Abraham Leo ², S. Thumilvannan ³

Affiliations
1 Department of Computer Science and Engineering, Kings Engineering College, IN
2 Department of Computer Science and Engineering, Kings Engineering College
3 Department of Computer Science and Engineering, Kings Engineering College, IN

Abstract

In surveillance, the need for enhanced image representation and rendering from low-resolution videos is paramount for effective analysis and decision-making. This research addresses the limitations of conventional methods in extracting meaningful information from low-resolution footage. The prevalent challenge lies in the compromised clarity and detail inherent in surveillance videos, hindering accurate identification and analysis of critical events. The ubiquity of surveillance cameras has led to an influx of low-resolution videos, limiting the efficacy of traditional image processing techniques. This research aims to bridge this gap by leveraging DenseNet, a densely connected convolutional neural network (CNN) known for its ability to capture intricate features. The DenseNet seeks to enhance the representation and subsequent rendering of images, transcending the constraints imposed by low resolutions. The network ability to capture intricate details will be harnessed to enhance image representation. Subsequent rendering techniques will be employed to reconstruct high-quality images for improved analysis. The results showcase promising advancements in image representation and rendering using DenseNet. The enhanced visual quality of surveillance images allows for more precise identification and analysis of events, demonstrating the potential impact of the proposed methodology on improving surveillance systems.

Keywords

Surveillance, Low-Resolution Videos, DenseNet, Image Representation, Rendering

Full Text

     

M. Parameswari ¹, B. Yuvaraj ¹, S. Thumilvannan ¹, E. Munuswamy ¹

Affiliations
1 Department of Computer Science and Engineering, Kings Engineering College, IN

Abstract

The Internet of Things (IoT) with advanced machine learning techniques presents significant potential for agricultural applications, particularly in the domain of crop disease detection. Paddy, a staple food for millions, is highly susceptible to various diseases that can drastically affect yield and quality. Early and accurate disease detection is crucial for effective management and mitigation. Traditional methods are often labor-intensive and less reliable, underscoring the need for automated, accurate, and scalable solutions. The primary challenge lies in developing a robust system capable of accurately identifying diseases in paddy leaves using IoT-collected data. This task is complicated by the variability in disease manifestation and environmental conditions, which can affect the quality and consistency of the collected data. Efficient feature extraction and classification techniques are essential to address these issues and ensure high accuracy. This study proposes a novel approach combining Particle Swarm Optimization (PSO) for feature extraction with Deep Belief Networks (DBNs) for classification. IoT devices capture highresolution images of paddy leaves, which are then processed in the cloud. PSO is employed to optimize the feature extraction process by selecting the most relevant features from the image data. These optimized features are fed into a DBN, which is trained to classify the images into healthy or diseased categories. The use of cloud computing ensures the scalability and computational efficiency of the system. The proposed method demonstrates significant improvements in accuracy and processing speed. The PSO-based feature extraction enhances the relevance of features, reducing the dimensionality and improving the DBN's performance. Experimental results show an accuracy rate of 96.3%, with a reduction in processing time by 35% compared to traditional methods. The system's precision and recall rates are 95.8% and 94.7%, respectively, highlighting its effectiveness in real-world applications.

Keywords

IoT, Crop Disease Detection, Paddy Leaves, Particle Swarm Optimization, Deep Belief Networks

Full Text