DETEKSI GERAK DENGAN MENGGUNAKAN BACKGROUND SUBTRACTION DAN GLOBAL THRESHOLDING (Studi Kasus: Universitas Muhammadiyah Purwokerto)

Digital image processing has now penetrated to almost all elements of digital life both research, health, and even to the realm of security. In the security field, the detection of object motion is crucial not only the recording media that plays a role but also the method used in the detection. In this study the author tries to combine background subtraction and global thresholding to detect object motion where the author chooses pedestrian motion at Universitas Muhammadiyah Purwokerto with a single pedestrian object type. While for recording media the author uses the camera of an Android phone with 1080p Full High Definition (FHD) resolution with a 60 fps framerate (frame per seconds). In this study the author divides the detection of the motion of objects into 2 backgrounds, including solid colored backgrounds and views of each applied with 4 values of global thresholding. From the results of the research conducted, the detection of pedestrian objects runs quite well and accurately on global thresholding values 150 on solid colored backgrounds with an average MSE value of 16.3462 and PNSR 13.0346. The end of this study is that motion detection can also be influenced by aspects such as the color of the object's object and the complexity of the background used.