STRAYFREE: ONLINE PLATFORM FOR STRAY DOG REPORTING WITH CNN-BASED AI MATCHING PICTURE

Abstract

Stray dogs pose a significant challenge in many communities, increasing risks to public safety and health. Traditional stray dog management methods, such as manual reporting and identification, are inefficient and slow, leading to delays in response and increasing incidents related to stray dogs. This study introduces StrayFree: Online Platform for Reporting Stray Dogs with CNN-Based AI Matching Picture, a web and mobile-based system designed to streamline stray dog reporting and identification. The platform integrates features such as dog registration, AI- powered image matching for dog identification, real-time stray dog reporting, and centralized data management to enhance efficiency. A descriptive-developmental methodology was employed using an Agile approach, involving iterative development, testing, and deployment. Technologies such as Python, TensorFlow, MySQL, and React Native were utilized to ensure scalability and usability. The evaluation involved Tuy residents, barangay officials, and Municipal Agriculture Office representatives, measuring system usability and satisfaction through the System Usability Scale (SUS). Results indicate a good usability rating (73.10 SUS score) and high satisfaction levels, particularly in dog registration (4.47), AI dog identification (4.32), and stray dog reporting (4.50). The CNN model achieved 86.65% accuracy for dog identification and 64.62% for breed classification. Findings suggest that StrayFree significantly improves stray dog management, enhances public safety, and promotes responsible pet ownership. Future recommendations include expanding the system’s coverage, improving AI accuracy, and integrating real-time tracking for better monitoring.

Keywords