Skripsi
IDENTIFIKASI JUMLAH KENDARAAN MENGGUNAKAN YOLO DAN PREDIKSI ARUS LALU LINTAS DENGAN MENERAPKAN ALGORITMA LSTM SERTA VISUALISASI HASIL BERBASIS WEBSITE PADA JALAN RAYA KOTA PALEMBANG
The annual increase in the volume of vehicles, without a corresponding improvement in road infrastructure, has led to detrimental traffic congestion and density. Therefore, this research aims to address this issue by developing a vehicle detection system using the You Only Look Once (YOLO) generation 8 algorithm and DeepSort architecture to count the number of vehicles. In addition to focusing on object detection, the study also considers the use of Long Short-term Memory (LSTM) method to assess road conditions as smooth, moderate, or congested. The dataset involves 3592 image files and 72 video files containing information about vehicles such as motorcycles and cars. However, the dataset model consists of five variables: motorcycles, cars, red lights, green lights, and zebra crossings. YOLOv8, generated from the image dataset, exhibits training and testing accuracies of 93.73%. This reflects excellent performance in detecting vehicle objects. Furthermore, the creation of the LSTM model for road condition detection yields an accuracy of 94.27%. Testing the video output in CSV format indicates that Monday mornings tend to be congested, while Wednesdays are relatively smooth. Meanwhile, on Fridays and Saturdays, the road conditions are generally moderate. The testing results are visualized through a locally-based PHP-programmed website in the form of graphs, facilitating a more interactive understanding and analysis of the research outcomes.
Inventory Code | Barcode | Call Number | Location | Status |
---|---|---|---|---|
2407000272 | T137966 | T1379662024 | Central Library (Referens) | Available but not for loan - Not for Loan |
No other version available