International Journal on Recent and Innovation Trends in Computing and Communication
Not a member yet
8613 research outputs found
Sort by
Unsupervised Machine Learning based Energy Efficient Routing for Mobile Ad-Hoc Networks
Get PDFMobile Ad-hoc Networks (MANETs) are temporary networks formed by a group of mobile hosts without the need for centralized administration or specific support services. Energy consumption is a critical issue in MANETs due to their reliance on limited battery resources. Reducing energy consumption is crucial for increasing network lifespan and throughput. Existing energy-saving techniques often fall short in their effectiveness. This research proposes a novel approach that combines a proactive MANET routing protocol with an energy-efficient strategy to address these limitations. The proposed solution considers both node mobility and energy levels in the routing process. Traditional AODV routing relies on flooding, which broadcasts RREQ packets to all nodes within the sender's range. This often leads to unnecessary retransmissions of RREQ and RREP packets, resulting in collisions and network congestion. To overcome this issue, we propose an optimized route discovery mechanism for AODV. The key idea is to leverage the K-means clustering algorithm to select the optimal cluster of nodes to forward RREQ packets instead of relying on broadcasting. This approach aims to alleviate network congestion and reduce end-to-end delay by minimizing unnecessary control packet transmissions
Fruit Detection and Classification using YOLO Models
Get PDFComputer Vision and Deep Learning techniques have become an advent in multiple domains like healthcare, Technology, as well as Agriculture . Computer vision techniques like object detection are being widely used in agriculture to reduce to efforts required and make agriculture a little more efficient for the farmers. The applications of deep learning in agriculture include leaf disease detection and weather forecasting, and the most advent applications include object detection to detect fruits, and vegetables which can be ensembled with robotics for automated yield production and harvesting. The proposed article describes one such application of fruit detection using various YOLO (You Only Look Once) models. The study encompasses four fruit classes namely Chiku, Mango, Mosambi, and Tomato. Models of Yolo V3, Yolo V4, and Yolo V8 were trained on a customized dataset collected from Indian farms and fruit gardens. The real time images images were collected, pre-processed, and annotated using online labeling tools. A total of 1200 images were used as a part of the complete training process. Basic preprocessing was performed on these images and possible inbuilt augmentation techniques supported by the above-mentioned models were used.Training is applied on custom dataset for all classes. In this experiment we have received the F1 score for YOLOv3(Chiku-82%.Mamgo-91%,Mosambi-87%,,Tomato-77%),YOLOv4(Chiku-89%.Mamgo-98%,Mosambi-95%,,Tomato-91%) and YOLOV8 (Chiku-90%.Mamgo-75%,Mosambi-82%,,Tomato-84%)models. In these models YOLOv4 with two layers gives the highest accuracy for all the classes
Advancements in Machine Learning for the Diagnosis of Chronic Kidney Disease
Get PDFChronic Kidney Disease (CKD) constitutes a significant global health issue, precipitating damage to the kidneys and stripping many individuals of their most productive years. Alarmingly, 40% of those affected by CKD remain oblivious to their condition, a stark contrast to many other diseases where early detection is more common. Unlike other conditions, CKD eludes cure unless identified promptly in its nascent stages. This research emphasizes the collection of critical indicators such as blood pressure and diabetes status to ascertain the presence of CKD in individuals. It proposes the employment of advanced machine learning techniques, including Random Forest, XGBoost, and Support Vector Machines, aiming to enhance early detection and thereby mitigate the disease's impact. Utilizing a CKD dataset, this study endeavors to predict the likelihood of CKD in individuals, offering a proactive approach to tackle this formidable health challenge
Analysis of Blockchain-Based Security Solutions for IoT Communication Networks
Get PDFIn recent years, the concept of blockchain technology has garnered substantial attention owing to its pivotal role as the underlying innovation powering cryptocurrencies such as Bitcoin. This surge in interest is particularly notable due to the manifold applications of blockchain technology across diverse domains, including but not limited to, bolstering the security landscape of the Internet of Things (IoT), fortifying the banking sector, optimizing industrial operations, and enhancing clinical establishments. Furthermore, the IoT paradigm has witnessed exponential growth in its adoption, primarily attributed to its seamless integration within smart homes and urban infrastructure projects on a global scale. However, a notable drawback within IoT lies in the inherent limitations of processing power, constrained data storage capacities, and limited network bandwidth of IoT devices. Because of these limitations, IoT devices stand more vulnerable to various forms of cyberattacks in comparison to their counterparts such as smartphones, tablets, or personal computers. This scholarly paper critically delves into the profound security challenges prevalent within the IoT ecosystem and meticulously examines the intricacies of addressing these challenges through the integration of blockchain technology. Additionally, this study identifies and elucidates certain dimensions that remain inadequately covered by existing blockchain implementations in IoT security contexts. Through this comprehensive exploration, the research aims to contribute to a deeper understanding of the potential and limitations of blockchain in fortifying IoT security, shedding light on unexplored avenues and underscoring the imperatives for future research and technological enhancements
Review Paper on Systematic Study of Leaf Disease Detection Using Accurate and Efficient ML Technique.
Get PDFIn recent years, plant diseases have posed significant threats to global food security and ecosystem stability. Timely detection and management of these diseases are imperative to mitigate their adverse effects. This paper introduces Foliage Guard, a novel smart plant leaf disease detector leveraging machine learning techniques for accurate and efficient disease identification. Leaves Guard employs state-of-the-art image processing algorithms to analyze leaf images captured using low-cost sensors or smartphones. The system utilizes a deep learning architecture trained on a diverse dataset of plant diseases to classify the health status of leaves accurately. Additionally, Foliage Guard incorporates real-time disease monitoring and alert mechanisms, enabling farmers and gardeners to take pro active measures against outbreaks. Through extensive experimentation and validation of various plant species, Foliage Guard demonstrates superior performance compared to existing approaches, with high accuracy and rapid processing times. The proposed system holds promise for revolutionizing plant disease management practices, offering a cost-effective and accessible solution for early disease detection and prevention in agriculture and horticulture sectors
Determination of Blast Disease Using SVM And ANN Classifiers
Get PDFThis paper is mainly industrialized to find the blast disease and reduce the crop defeat and hence increase the paddy cultivation production in an effective manner. In modern farming field, pest and disease identification is a major role of paddy cultivation. Image classification by the use of deep convolutional neural networks of training and methodology used the facilitate a quick and easy system implementation. Pests and diseases are a threat to paddy production, especially in India, but identification remains to be a challenge in massive scale and automatically. The results show that we can effectively detect and recognize the paddy diseases and pests including healthy plant class using classifiers, with the best accuracy of 91%. The significantly high success rate makes the model a really useful advisory or early warning tool, and an approach that would be further expanded to support an unified paddy plant disease identification system to work in real cultivation conditions
Optical Flow Approach for Real-Time Crowd Activity Identification Using Segmentation
Get PDFThis research introduces an optical flow approach for real-time crowd activity identification using segmentation methods. The methodology focuses on utilizing the optical flow field to analyze motion patterns within crowd scenes, thereby segmenting the scene into coherent motion regions for activity identification. Through clustering of optical flow vectors, distinct motion regions are delineated, facilitating the identification of various crowd activities such as walking, running, and gathering.Efficient algorithms for optical flow computation and segmentation are implemented to ensure real-time performance. The research is validated through implementation and testing on diverse crowd scenes, demonstrating its efficacy in accurately identifying different crowd activities in real time. Experimental results showcase the superiority of the optical flow-based segmentation method over traditional techniques in terms of accuracy and computational efficiency, thus presenting a promising solution for real-time crowd activity identification systems
Waymark in the Depths: Baseband Signal Transmission and OFDM in Underwater Acoustic Propagation Channel Models
Get PDFIn the intricate environment of underwater acoustic propagation, establishing reliable communication channels stands as a formidable challenge, primarily due to the medium's inherent properties, such as high path loss, multipath propagation, and time-varying channel characteristics. "Waymark in the Depths: Baseband Signal Transmission and OFDM in Underwater Acoustic Propagation Channel Models" presents an innovative exploration into enhancing underwater communication systems by leveraging advanced signal processing techniques and channel modeling strategies. At the core of this research lies the integration of Orthogonal Frequency Division Multiplexing (OFDM) with baseband signal transmission, aiming to mitigate the detrimental effects of the underwater acoustic environment on signal integrity and throughput. By dissecting the acoustic channel's unique attributes, the study devises a comprehensive channel model that encapsulates the dynamic nature of underwater acoustics, including the impact of temperature, salinity, and pressure on sound speed and signal dispersion. This model serves as a waymark, guiding the development of tailored OFDM techniques that are optimized for the underwater medium, focusing on maximizing spectral efficiency and minimizing error rates. The research meticulously examines the interplay between baseband signal processing and OFDM in this context, illustrating how their synergistic application can overcome the bandwidth limitations and frequency-selective fading characteristic of underwater channels. Through extensive simulation and experimental validation, the study demonstrates the feasibility of achieving high-speed, reliable underwater communication, highlighting significant improvements in data rates and link stability. Furthermore, the research delves into adaptive modulation schemes and coding strategies, optimized for the derived channel model, to bolster the robustness of the communication link against the unpredictable underwater environment. This pioneering work not only sheds light on the complexities of underwater acoustic signal transmission but also charts a path forward for the next generation of underwater communication systems. By pushing the boundaries of current technological capabilities and offering a solid theoretical foundation, this research contributes significantly to the field of underwater acoustics and opens new horizons for marine exploration, environmental monitoring, and submarine communication networks. Through its comprehensive analysis and innovative approaches, "Waymark in the Depths" not only addresses the technical challenges of underwater signal transmission but also lays down a crucial waymark for future endeavors in the uncharted territories of the ocean's depths
Emoticon Generation, Expression Recognition, and Gender Classification Using Deep Learning in Real-Time
Get PDFImages play an increasingly important role in identifying a person's gender and emotional state in today's digital environment, but there are still methodological hurdles to overcome. Image processing utilizing deep learning algorithms is the way to go. Our study's overarching goal is to find ways to bridge the communication gap through the use of emoticons based on the emotions conveyed in photographs and snapshots. We have utilized the Keras framework to implement a deep learning algorithm called a Convolutional Neural Network (CNN) and evaluated it using Tensor Flow to predict gender. The goal is to create a new dataset of pictures free of noise and then utilize those images as inputs to a convolutional neural network (CNN). The algorithm's result is supposed to be more trustworthy gender identification based on increased accuracy. We have implemented an LSTM-RNN (Long short-term memory recurrent neural network) for emotion identification and facial expression detection. Feature selection is the most crucial step since it will ultimately aid in emoticon generation