Crash kita: implementing convolutional neural network for vehicular accident detection and assistance on tirona highway, bacoor city, cavite/
Cruz, Wency Gabriel O.
Crash kita: implementing convolutional neural network for vehicular accident detection and assistance on tirona highway, bacoor city, cavite/ Wency Gabriel O. Cruz, Joshua Louise B. Margallo, Maverick Rayne A. Ramos, Carl C. Senaris, and Sandryl A. Torres.-- - Manila: Technological University of the Philippines, 2025. - x, 140pages: 29cm.
Bachelor's thesis
College of Science.--
Includes bibliographic references and index.
In the Philippines, traffic accidents are still a major issue, particularly along the
Tirona Highway in Bacoor City, Cavite. The goal of this study, "CRASH KITA:
Implementing Convolutional Neural Network for Vehicular Accident Detection and
Assistance," is to develop a model that uses computer vision and IP cameras to
automatically detect accidents. The researchers used two pre-trained YOLOv11 models to
recognize different types of accidents. One focused on car-to-car collisions and reached
around 84% mean average precision and 75% accuracy, while the other model included
motorcycles with lower accuracy at 70%. The system includes a website made with PHP,
Tailwind, and Bootstrap, where barangay authorities can view, confirm, manage
accidents, and create reports. Testing showed that the system works well, but with
limitations depending on the camera angle. Overall, “CRASH KITA” shows that using
machine learning can help improve road safety monitoring by speeding up accident
detection and response.
Computer vision
Road safety
IP cameras
BTH QA 76 / C78 2025
Crash kita: implementing convolutional neural network for vehicular accident detection and assistance on tirona highway, bacoor city, cavite/ Wency Gabriel O. Cruz, Joshua Louise B. Margallo, Maverick Rayne A. Ramos, Carl C. Senaris, and Sandryl A. Torres.-- - Manila: Technological University of the Philippines, 2025. - x, 140pages: 29cm.
Bachelor's thesis
College of Science.--
Includes bibliographic references and index.
In the Philippines, traffic accidents are still a major issue, particularly along the
Tirona Highway in Bacoor City, Cavite. The goal of this study, "CRASH KITA:
Implementing Convolutional Neural Network for Vehicular Accident Detection and
Assistance," is to develop a model that uses computer vision and IP cameras to
automatically detect accidents. The researchers used two pre-trained YOLOv11 models to
recognize different types of accidents. One focused on car-to-car collisions and reached
around 84% mean average precision and 75% accuracy, while the other model included
motorcycles with lower accuracy at 70%. The system includes a website made with PHP,
Tailwind, and Bootstrap, where barangay authorities can view, confirm, manage
accidents, and create reports. Testing showed that the system works well, but with
limitations depending on the camera angle. Overall, “CRASH KITA” shows that using
machine learning can help improve road safety monitoring by speeding up accident
detection and response.
Computer vision
Road safety
IP cameras
BTH QA 76 / C78 2025