Volume 13 | Issue 7

 Abstract

This study explores the implementation of an Adaptive In-Loop Filter (AILF) for High Efficiency Video Coding (HEVC) utilizing deep learning technique. The need for effective compression techniques that preserve excellent visual quality has grown as video content continues to spread. Even though the traditional in-loop filters perform well, they frequently have difficulties maximizing performance in a variety of video scenarios and environments. According to the properties of the video being analyzed, this study suggests an AILF that includes Convolutional Gated Recurrent Unit (ConvGRU), a type of Recurrent Neural Network typically involves enhancing reconstructed frames by exploiting temporal dependencies across frames. In addition to enhancing reconstructed frames, the AILF performs better than traditional techniques in terms of Structural Similarity Index (SSIM) and Peak Signal-to-Noise Ratio (PSNR), indicating its potential for practical uses in broadcasting and video streaming. By demonstrating how well deep learning techniques can be integrated into video processing tasks, this work adds to the continuous developments in video coding technology.

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

Adaptive in-loop filter, Deep learning techniques, High Efficiency Video Coding, Convolutional Gated Recurrent Unit, Neural Network.

  License

Creative Commons Attribution 4.0 and The Open Definition

 Abstract

This paper presents the design and development of an Internet of Things based solution for urban environmental monitoring and pollution hotspot detection. The system integrates with a Espressif Systems 32-WROOM microcontroller for sensor data acquisition, while the Espressif Systems 32 Wi-Fi module processes the data and sends it to a central server, where Geographic Information System mapping is used to visualize pollution hotspots in real time. The Long Range Radio Frequency RF module extends the communication range in urban areas with limited cellular coverage. Geographic Information System mapping helps visualize pollution levels across urban areas, facilitating the identification and management of pollution hotspots. To enhance pollution mitigation, the system integrates an Active Carbon Filter, which helps reduce airborne pollutants before data collection. The filter is placed near gas sensors (MQ-135, MQ-7, MQ-8) to compare pre-filtered and post-filtered air quality, providing insights into filtration efficiency. It actively removes harmful gases such as carbon monoxide, Nitrogen Oxides, and Volatile Organic Compounds, supporting sustainable urban management. This system supports data-driven decision-making and timely interventions, contributing to sustainable urban planning and public health initiatives. The system aims to offer valuable insights into pollution trends, aiding in the identification of critical areas for intervention. Additionally, the use of a low-power, low-cost microcontroller makes it suitable for large-scale deployment in urban environments.

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

Internet of Things (IoT), Long Range radio Frequency (LoRa RF), ESP32 Wi-Fi module, Metal Oxide (MQ), carbon monoxide (CO), Nitrogen Oxides (NOx), Volatile Organic Compounds (VOCs), Geographic Information System (GIS).

  License

Creative Commons Attribution 4.0 and The Open Definition

 Abstract

IoT-Based Intelligent Trolley System enhances shopping by integrating IoT and embedded systems to automate product selection and billing. It eliminates long checkout lines by enabling real-time billing within the trolley. The system includes an Arduino Uno, Radio Frequency Identification based product scanning, and a display for product details. New customers enter their mobile numbers via a keypad to receive offers through Short Message Service, while regular customers use Radio Frequency Identification membership cards for personalized recommendations. Products are added or removed using Radio Frequency Identification taps. After shopping, the customer presses a finish button to trigger billing. The total bill is sent via Bluetooth to a thermal printer for receipt generation and via Wireless Fidelity through Node Mirco Controller Unit to the store owner's Personal Computer. A Quick Response code inside the store provides access to a website for browsing products, viewing discounts, and tracking expenses. This system improves efficiency, reduces manual errors, and enhances the shopping experience.

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

IoT, Radio Frequency Identification, Real-time Billing, Bluetooth, Wireless Fidelity, Automated Checkout.

  License

Creative Commons Attribution 4.0 and The Open Definition

 Abstract

Sewage block detection and removal presents an automated system for detecting and removing sewage blockages using sensors and microcontroller-based control. Ultrasonic and water flow sensors monitor water levels and flow rates in real time. When abnormal readings indicate a blockage, a relay-controlled pump is activated to flush the system. The design aims to reduce manual intervention, improve sanitation, and prevent overflow in urban and residential settings. The system is cost-effective, scalable, and suitable for smart city applications.

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

IoT, Sewage management, Real-time monitoring, Automation, Blockage removal

  License

Creative Commons Attribution 4.0 and The Open Definition

 Abstract

Low-Density Parity-Check (LDPC) codes are a key error correction technique in communication systems, ensuring reliable data transmission over noisy wireless channels. Traditional LDPC decoders, such as the Min-Sum algorithm, use fixed normalization (?) factors to approximate belief propagation. However, these fixed parameters do not adapt to varying channel conditions, leading to suboptimal performance in real-world scenarios. This project introduces a gradient descent-based approach to dynamically optimize ? based on estimated Signal-to-Noise Ratio and Rayleigh fading conditions. The system predicts the optimal values for ? in real time, improving decoding efficiency and error correction performance. These predicted parameters are then used in the Min-Sum decoding process, reducing the number of iterations required for convergence and minimizing the Bit Error Rate. The proposed approach enhances the adaptability of LDPC decoding, significantly improves error correction performance, reduces decoding time, and adapts effectively to varying channel conditions.

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

LDPC codes, error correction, communication, Min-Sum algorithm, normalization factor, offset factor, belief propagation, Rayleigh fading, SNR, gradient descent optimization.

  License

Creative Commons Attribution 4.0 and The Open Definition

 Abstract

One of the major causes of irreversible blindness is Glaucoma. It causes progressive damage to the optic nerve, often without noticeable symptoms in the early stages. Detection of this condition is vital for averting vision loss. Recent technological advancements in medical imaging technologies coupled with new and improved computational methods have enabled significant progress in glaucoma diagnosis. This literature review examines the evolution of automated glaucoma detection, with a focus on the role of image preprocessing, feature extraction, and the use of machine learning (ML) along with deep learning (DL) techniques. The review highlights essential preprocessing methods to enhance image quality, such as contrast enhancement and noise reduction, which are critical for accurate analysis of fundus and OCT images. Additionally, it explores various feature extraction approaches that bridge raw image data to meaningful clinical insights. This comprehensive review also provides an overall picture of different ML and DL models employed to detect glaucoma, evaluating their strengths, limitations, and performance metrics. Furthermore, it addresses the challenges faced in the field, such as dataset imbalance, the need for diverse and high-quality datasets, and the integration of these automated systems into clinical practice. The paper concludes by discussing future directions for research, including the potential of hybrid models, multimodal frameworks, and improved interpretability in advancing glaucoma detection and management.

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

Glaucoma, Retinal Imaging, Optical Coherence Tomography (OCT), Fundus Images.

  License

Creative Commons Attribution 4.0 and The Open Definition

 Abstract

This paper gives a real-time landslide detection and traffic automation system based on ADXL sensors, IR sensors, and rain sensors. The system gives real-time alerts using an LCD display and manages traffic by motorized gates. Road safety and accident avoidance are enhanced with wireless communication based on Zigbee and emergency notification based on GSM.

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

Landslide detection, traffic automation, ADXL sensor, IR sensor, rain sensor, ESP 32, Zigbee communication, GSM module, road safety, real-time monitoring.

  License

Creative Commons Attribution 4.0 and The Open Definition

 Abstract

The IoT-based Dam Automation System proposed combines microcontrollers, sensors, and image processing to improve the efficiency and safety of dam operation. With ESP32 microcontrollers, ultrasonic sensors, turbidity sensors, rain sensors, and a GSM module, the system tracks important parameters such as water levels, water quality, and structural condition. Camera-based image processing identifies cracks or damages in the dam structures and sends real-time notifications to concerned authorities. The system also has an automatic control mechanism that operates to manage the dam gates according to water levels in order to avert overflow and optimize water allocation. This solution minimizes dependency on manual adjustment, enhances preparedness for disaster, and presents a consistent, real-time alternative for more effective and safer operation of contemporary dam infrastructure.

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

Automation, Camera, ESP32, water level, monitoring, crack, dam.

  License

Creative Commons Attribution 4.0 and The Open Definition

 Abstract

Cleaning staircases is tedious and labor-intensive. This work presents an autonomous staircase-cleaning robot. It efficiently traverses and cleans stairs using a NEMA17 stepper motor for climbing. Rear wheels are powered by DC motors, while a front-side DC motor enables rotation. The cleaning system includes spinning brushes, suction, and optional water sprays. 'Proximity' and 'edge' sensors ensure safe operation. A microcontroller-based system controls movement and scrubbing. This robot speeds up cleaning and reduces manual effort. It enhances cleanliness in various environments. Suitable for homes, offices, and public spaces. A smart, efficient cleaning solution

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

Autonomous, DC motors, edge sensors, intelligent cleaner, microcontroller, NEMA17 stepper motor, spinning brushes, staircase-cleaning robot, suction system, water sprays

  License

Creative Commons Attribution 4.0 and The Open Definition

 Abstract

An amphibious hovercraft is a highly adaptable vehicle capable of traversing both land and water by riding on a cushion of air produced by powerful fans. This unique mode of operation allows it to glide effortlessly over diverse surfaces such as water, mud, sand, and ice, making it particularly valuable in areas with difficult or unstable terrain. Its versatility makes it ideal for use in flood zones, wetlands, and remote regions where conventional vehicles often face limitations. The hovercraft's seamless transition between land and water also makes it an effective tool for rapid emergency response, including rescue operations and disaster relief efforts. Beyond emergency services, hovercrafts are utilized in military operations, passenger and cargo transport, and recreational activities. However, to achieve broader adoption, challenges such as high fuel consumption, noise levels, and complex maintenance requirements must be addressed. Modern designs now incorporate advanced features like GPS navigation, improved propulsion systems, and robust construction materials to boost performance, efficiency, and reliability.

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

Skirt Design, Hybrid Propulsion, Durable Materials, Propulsion systems, sustainable technology

  License

Creative Commons Attribution 4.0 and The Open Definition

 Abstract

This project is related to the invention of a vehicle security and monitoring system that integrates geo-fencing and real-time tracking. Through the use of Arduino Mega 2560, GPS, GSM, and sensors, the system offers theft detection, accident monitoring, and secure access control. In geo- fencing, boundary alerts are obtained, and, whereas, load monitoring is unauthorized use. The system fuses various security technologies executing an excellent and cost-efficient solution for both fleet management and the use of passive theft prevention, along with real-time vehicle tracking

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

Geo-fencing, Vehicle Security, Theft Detection, Accident Monitoring, Boundary Alerts, Load Monitoring

  License

Creative Commons Attribution 4.0 and The Open Definition

 Abstract

This research presents an autonomous surveillance rover for defense applications designed to enhance real-time enemy detection and situational awareness. The system is powered by a Raspberry Pi, enabling centralized control and processing of data from various sensors. Ultrasonic sensors ensure obstacle detection, while inductive proximity sensors identify landmines and explosive devices. A camera module provides real-time video streaming, which is analyzed using image processing algorithms to classify individuals as authorized or threats. When a threat is identified, the system triggers a laser module to simulate a defensive response. The rover autonomously navigates its environment, continuously scanning for hazards and relaying real-time alerts to defense personnel via a communication module. This integration of IoT, artificial intelligence, and robotics makes the rover a reliable and efficient solution for modern defense challenges. The project highlights its potential to enhance security, reduce human risks, and adapt to evolving operational demands.

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

Autonomous surveillance, enemy detection, real-time monitoring, Raspberry Pi, obstacle detection, image processing, IoT, defense robotics.

  License

Creative Commons Attribution 4.0 and The Open Definition

 Abstract

The Wildlife Observation Robot represents an innovative solution for automated forest surveillance and protection. This autonomous system integrates advanced computer vision capabilities with robust hardware components, all controlled by a Raspberry Pi microcomputer. The robot continuously patrols forest areas, utilizing a rotating camera system and Ultrasonic sensors for comprehensive environmental monitoring. Through OpenCV-based image processing, it can detect both wildlife presence and potential fire outbreaks in real-time. Upon detection, the system immediately alerts relevant authorities via Telegram messaging, providing crucial information including precise location coordinates and photographic evidence. This implementation significantly enhances forest management capabilities while reducing human intervention in potentially dangerous situations.

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

Autonomous Wildlife Monitoring, Smart Surveillance System, Automated Alert System

  License

Creative Commons Attribution 4.0 and The Open Definition

 Abstract

The usage of chemical herbicides is common in the labor-intensive agricultural process of weeding, which can be hazardous to both human health and the environment. The YOLO (You Only Look Once) model is an autonomous weed identification robot that uses computer vision and machine learning to overcome these obstacles because of the robot's accurate weed identification and categorization, less chemical pesticide is used, labor costs are decreased, and sustainable farming is promoted. The method involves designing and building a prototype robot, training and testing its algorithms, and evaluating how well it performs in real-world situations. Anticipated outcomes include enhanced weed management, cost reductions, higher productivity, sustainability, and advancements in agricultural technology. This creative method combines automation and AI-driven decision-making to revolutionize conventional farming methods.

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

IoT in Agriculture, Crop Health Monitoring, Autonomous Weed Identification, Agricultural Robotics, and Weed Management.

  License

Creative Commons Attribution 4.0 and The Open Definition

 Abstract

As the demand for eco-friendly solutions grows, the EcoNav Drone offers a smart and sustainable approach to both aerial navigation and plant health monitoring. Equipped with GPS for navigation and a camera for real-time data collection, the drone operates with manual control. The SpeedyBee F405V3 flight controller ensures stability and adaptability in various environments. With intelligent route planning and adaptive flight control, the drone operates efficiently while consuming less power. In addition, early detection of plant diseases is vital for maintaining crop health and boosting agricultural productivity, making the EcoNav Drone a valuable tool for sustainable farming practices.

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

Adaptive flight control, Aerial data collection, Autonomous navigation, Eco-friendly drones, Energy-efficient flight, Environmental monitoring, GPS-based navigation, Low-power UAV, Optimized route planning, Sustainable drone technology.

  License

Creative Commons Attribution 4.0 and The Open Definition

 Abstract

The proposed Smart Pepper Spray is an advanced self-defense device that provides personal security through the integration of modern technology in contrast to traditional pepper spray the device includes GPS tracking and a high-resolution camera which monitors location surveillance in real time and provides visual evidence of emergency visual evidence the device is equipped with a sensor that provides GPS - as sow which GPS-as sow uses the amount spent record and store integrated cameras and film material from incidents that may serve as important evidence for film enforcement agencies compact and user-friendly design ensures portability and user-friendly while operating the system for long-term reliability with a rechargeable battery.

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

Snart Pepper Spray, Microcontroller ESP8266, Push Button, Servomotor, Bluetooth Module HC-05, Web Camera, Charge Controller Module.

  License

Creative Commons Attribution 4.0 and The Open Definition

 Abstract

This project presents a Traffic Rules Violation Detection System using Raspberry Pi, Pi Camera, and a sound sensor. It detects unauthorized license plates, exhaust modifications, and dangerous stunts like wheelies in real time. Deep learning and computer vision techniques are used for video and audio analysis. The system automates violation detection, reducing human effort and enhancing traffic rule enforcement

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

HSRP detection, Number plate recognition, vehicle violation detection ,image processing.

  License

Creative Commons Attribution 4.0 and The Open Definition

 Abstract

A surveillance camera is an essential security device used for monitoring and recording activities in a specific area. It is commonly used in various settings such as homes, businesses, public spaces, and industrial sites to deter crime, monitor behavior and ensure safety. Surveillance cameras are typically connected to a system for remote viewing and recording, allowing users to observe live footage or review past events. AI-powered surveillance cameras represent a significant advancement in security technology, integrating artificial intelligence and machine learning algorithms to provide more intelligent, automated, and accurate surveillance. Unlike traditional surveillance cameras that simply capture and record footage, AI-powered cameras have the ability to analyze data in real-time, recognize patterns, and make decisions based on predefined criteria. This leads to enhanced security, greater efficiency, and faster response times. AI-powered cameras can automatically identify and track objects or people within their field of view. Using machine learning and computer vision algorithms, these cameras can differentiate between humans, vehicles, animals, and other objects. They can also detect specific behaviors or events such as Intrusion detection, Loitering detection etc.

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

Surveillance camera, Fire detection, scream detection, Arduino, Emergency services, smart security, IOT based , smart surveillance system.

  License

Creative Commons Attribution 4.0 and The Open Definition

 Abstract

Workers monitoring and safety assurance bot in oil refinery using esp32 CAM introduces an intelligent surveillance and safety monitoring system using a mobile robot built on ESP32 architecture. It features real-time face recognition, environmental monitoring, and emergency response capabilities, controlled via a mobile application with automated alert systems. Designed to enhance workplace safety, particularly in hazardous environments like oil refineries, the system detects hazardous gases, temperature fluctuations, and motion anomalies, providing live video feeds and alerts to a central hub. Leveraging low-cost IoT technology, it aims to reduce workplace accidents, improve incident response times, and ensure compliance with safety regulations, making it a scalable and adaptable solution for industrial safety.

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

Workers Monitoring, Safety Assurance, ESP32-CAM, IoT-Based Surveillance, Face Recognition, Environmental Monitoring, Industrial Safety, Real-Time Alerts, Wireless Communication, Hazard Detection, Remote Monitoring, Autonomous Surveillance Bot.

  License

Creative Commons Attribution 4.0 and The Open Definition

 Abstract

This project aims to develop a heliotropic system that enhance the performance of panels. It achieves this by using Light Dependent Resistor (LDR) sensors to detect sunlight intensity. Since the orientation of panels significantly impacts the amount of energy they collect, this system incorporates four LDR sensors positioned in different directions--North, South, East, and West. These sensors monitors sunlight levels from various angles, ensuring optimal positioning for greater energy absorption. The goal here is to design a heliotropic tracker that improves the panels performance using light sensors called Light Dependent Resistors (LDRs). It's well-known that light travels in straight lines from its source. In solar power systems, the alignment of solar panels with the setting sun is vital for maximizing energy capture. To achieve this, the project uses four LDR sensors arranged in a cross configuration, each facing North, South, East, and West. This is accomplished with the help of three servo motors that allow the solar panels to rotate. These motors control the horizontal (East-West) and vertical (North-South) movements, as well as the tilt of the system panels in relation to the sun's position. An Arduino microcontroller will manage the servo motors, directing the movement of the system panels based on the data received from the LDR sensors.

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

Component Solar Tracking System, Arduino Nano, Light Dependent Resistors (LDRs), Servo Motors, Solar Panel Orientation, Maximum Energy Absorption, Sunlight Intensity Detection, Buck Converter, Power Regulation, Automatic Sun Tracking, Real-Time Panel Adjustment, Renewable Energy Optimization, Dual-Axis Tracking, Efficiency Enhancement, Smart Solar Technology.

  License

Creative Commons Attribution 4.0 and The Open Definition