A mobile cloud solution for japanese koi recognition and recommendation system

Tan, Wei Thong (2024) A mobile cloud solution for japanese koi recognition and recommendation system. Final Year Project, UTAR.

Preview

PDF Download (6Mb) | Preview

Abstract

This project introduces a mobile cloud solution for Japanese Koi recognition and recommendation, designed to address significant challenges in the identification and grading of these captivating fish. The current lack of user-friendly technology for recognizing various types of koi fish, coupled with the time-consuming nature of learning to identify them, motivated the development of this innovative system. Additionally, the high market and business value of koi fish, often leading to fraud in purchases, emphasizes the need for a reliable recognition tool in the koi-fish-keeping community. The proposed system leverages Rapid Application Development (RAD) phased development methodology, encompassing planning, analysis, design, and implementation. Notably, it aims to bridge the learning gap for new enthusiasts by providing a convenient and accessible mobile platform for koi fish identification. The project utilizes cutting-edge technology, including cloud computing and Artificial Intelligence, to develop a system capable of accurately identifying koi fish breeds. A key feature of this system is real-time koi detection, allowing users to capture and recognize koi fish instantly through their mobile devices. The motivation behind this project extends beyond academic interests, addressing the practical needs of koi fish enthusiasts. The mobile application, designed for ease of use, allows users to upload images or capture live photos of koi fish for instant recognition. This not only streamlines the learning process but also promotes healthy koi trading by helping users confidently identify and differentiate various grades of Japanese koi. To enhance the learning experience, the system integrates a Chatbot, a large language model, providing users with interactive guidance and a dynamic educational environment. Furthermore, the system contributes to fraud prevention in koi fish trading by ensuring a confident environment for identifying different species and grades. By utilizing a diverse collection of koi fish datasets for training, the system employs cloud computing and AI to store and process information efficiently. The trained model, stored in the cloud, analyzes and predicts koi fish images uploaded or captured in real-time by users, offering detailed information such as type, size, and background. In conclusion, this project's contributions lie in reducing the learning gap for new koi enthusiasts, promoting healthy koi trading practices, and providing a comprehensive, user-friendly platform for recognizing the value of Japanese Koi.

Actions (login required)