Journal of Advanced Research in Computer Graphics and Multimedia Technology

Unveiling the lost city: Satellite-based detection of underwater ruins in Dwarka

Fri, 03 Oct 2025 00:00:00 +0000

The search for underwater archaeological sites poses a serious challenge because of the limitations of conventional exploration methods such as sonar sounding and underwater excavations. This research uses remote sensing to investigate the potential remains of ancient Dwarka, an historical and mythological site believed to be underwater in the Arabian Sea. Utilizing the computation power of Google Earth Engine (GEE), we process Sentinel-2 satellite imagery in order to detect underwater anomalies via the Normalized Difference Water Index (NDWI). Employing a threshold-based masking approach, we are able to separate high-NDWI regions that could be indicative of submerged structures from natural seabed variability. Our method integrates cloud-based geospatial analysis with interactive visualisation tools like geemap and Folium, allowing dynamic exploration of underwater morphology. This technique offers a scalable, low-cost, and non-destructive means of alternative marine archaeology that allows rapid appraisal of possible archaeological sites. The findings of this study contribute to the growing discipline of satellite-based underwater heritage discovery, demonstrating the revolutionary capability of Earth Observation (EO) technology in unearthing lost civilizations. Future work will focus on developing more advanced spectral analysis methods and using multi-source data sets to enhance the precision of underwater structure identification.

Behaviour Prediction of Shrimp using trajectory analysis and their validation with deployed IoT Sensor

Vinod Kumar Yadav, Rishik Yadav — Fri, 03 Oct 2025 00:00:00 +0000

This study presents a comprehensive end-to-end pipeline for real-time monitoring and behaviour prediction of shrimp locomotion in variable environmental conditions, integrating state-of-the-art deep learning, computer vision, and signal processing methodologies. The framework combines an enhanced YOLOv8 detection system with Deep SORT tracking algorithms and implements a sophisticated hybrid trajectory denoising approach utilising Savitzky-Golay filtering, cubic spline interpolation, and Gaussian smoothing. Applied to a custom-annotated dataset of 789 underwater shrimp images, the detection model achieved 0.74 precision, 0.856 recall, and 0.794 mAP@0.5. Advanced trajectory analysis techniques enabled 3D visualisation and directional behaviour quantification under four distinct environmental regimes combining pH variations (5.4.6.8) and temperature conditions (33°C, 35°C). A comprehensive comparative analysis with recent advances in underwater object detection, particularly YOLOv8-CPG architectures incorporating Compact Inverted Blocks (CIB), Partial Self-Attention (PSA), and Gold-YOLO feature fusion mechanisms, demonstrates potential performance improvements of 1-3% mAP through architectural optimisation. The methodology's integration with precision aquaculture monitoring systems, IoT sensor networks, and real-time behavioural alerting mechanisms positions it as a critical tool for sustainable aquaculture management, environmental stress detection, and automated welfare assessment in intensive farming operations.

Utilization of IoTs and AI to Modernize Academic Libraries

Tejas Shah, Mahesh K. Solanki, Paresh M Dholakia — Fri, 03 Oct 2025 00:00:00 +0000

Academic libraries have long been considered the cornerstone of higher education institutions, supporting learning, teaching, and research. With the onset of digital transformation and Industry 4.0 technologies, libraries are under increasing pressure to reinvent themselves to remain relevant to the modern academic community. Two of the most powerful technologies influencing this change are the Internet of Things (IoT) and Artificial Intelligence (AI). The integration of IoT devices with AI systems has created opportunities to design “smart libraries” that provide efficient resource management, enhanced security, personalised services, data-driven decision-making, and enriched user experiences. This paper provides an in-depth exploration of how IoT and AI can be applied to modernising academic libraries while also addressing challenges, case studies, and future prospects.

Smart Automated Warehouse System for E-Commerce Order Packaging Model

Nainesh Nageshree , Mansi Modi, Yash Akvaliya, Hiren Bhat — Sat, 11 Oct 2025 00:00:00 +0000

The rapid growth of e-commerce has greatly increased the demand for rapid, accurate and contact-free order supply. To meet these developed expectations, the automated warehouse system has emerged as a practical solution to increase operational efficiency. The project introduces a smart, automated warehouse system that basically integrates real-time web communication with physical automation to customise the packaging process.

Air Writing App: Leveraging Machine Learning and Fine-Tuning for Handwriting Recognition

Heena, Sandeep Ranjan — Wed, 21 Jan 2026 00:00:00 +0000

Handwriting recognition systems have evolved into a sophisticated intersection of artificial intelligence, deep learning, and optimisation. These systems aim to translate handwritten text into machine-readable formats, enabling applications in education, healthcare, banking, and beyond. However, achieving high accuracy, speed, and computational efficiency remains a significant challenge, especially when dealing with diverse handwriting styles and real-time requirements. Optimisation plays a critical role in improving the performance of handwriting recognition models. By fine-tuning hyperparameters, leveraging transfer learning, and optimising algorithms, developers can create systems that meet the demands of accuracy and scalability. This paper explores the
methodologies and techniques used to enhance handwriting recognition systems, focusing on hyperparameter tuning, real-time performance optimisation, and transfer learning.