Deteksi Api pada Video dengan Gaussian Mixture Model Untuk Deteksi Gerakan Dan Segmentasi Warna Api Dalam Ruang Warna YCbCr

Fire is a disaster that can endanger lives and cause property loss. The solution to detect fire commonly used today use sensor. sensors be together with surveillance cameras (CCTV) which are now being installed in many office buildings. This study tries build model for detecting video digital image processing approach using the Gaussian Mixture Model motion detection color segmentation YCbCr space. then tested metrics accuracy, precision, recall, speed. dataset form of videos small, medium, large sizes, only have fire-like objects. test results show algorithm able when size not too small or position close enough camera. For resolution 800x600 framerate 30 fps, it achieve 66.89% 73.77% 66.66% recall. performance during day relatively better than at night. Algorithm speed slow implemented real-time