DETEKSI DAN KLASIFIKASI KENDARAAN DARI HASIL VIDEO UDARA MENGGUNAKAN METODE GAUSSIAN MIXTURE MODEL
Daftar Isi:
- Data about traffic has an important role in urban planning, such as road planning. Therefore we need an algorithm that can detect and classify vehicles through video results. In this final project the technique of recording from a vehicle is explained and saving it in the form of a video file (*. Mp4), then filtering the vehicle using the Gaussian Mixture Model (GMM) with 2 types of distribution, the Bacgkround distribution and the Foreground distribution. The form of the Foreground distribution will be filtered using morphological operations, namely erosion operations and segmented using Bit Large Object (BLOB) to get the pixel area of the vehicle. BLOB results from these vehicles will be used as data for the classification process of vehicles in the form of motorcycles and cars. Experiments show the highest accuracy rate of success in the detection and classification of rear view vehicle video samples is 88.9% and the highest accuracy rate of success in detection and classification of front view vehicle video samples is 88.9%