Deteksi dan Penggolongan Kendaraan dengan Kalman Filter dan Model Gaussian di Jalan Tol

Monitoring systems are widely implemented in various sectors aimed at improving the security and productivity aspects. The research aims to detect moving objects in the form of video file tipefile (* .avi) 640x480 resolution and image class according to pixel area. Moving objects are given in the Region of Interest path for easy detection. Detection on moving objects using methods of Kalman filter and gaussian mixture model. There are two types of distribution, the distribution of Background and Foreground. The form of the Foreground distribution is filtered using Bit Large Object segmentation to obtain the dimensions of the vehicle and morphological operations. The feature extraction results from the vehicle are used for vehicle classification based on pixel dimension. Segmentation results are used by Kalman Filter to calculate the tracking of moving object positions. If the Bit Large Object segmentation is not found moving object, then it is continued on the next frame. The final results of system detection are calculated using Positive True validation, True Negative, False Positive, and False Negative by looking for the sensitivity and specificity of each morning, day and night conditions