Measuring the distance of crowd movement objects to the camera using a stereo camera calibrated with object segmentation histogram of oriented gradient

Bintang eka putera (Gunadarma University, Indonesia)
Singgih Jatmiko (Gunadarma University, Indonesia)
Ary Bima Kurniawan (Gunadarma University, Indonesia)

Jurnal Mantik
Vol. 7 No. 1 (2023): May: Manajemen, Teknologi Informatika dan Komunikasi (Mantik)

Measuring the distance of objects to human objects is currently under development. In its development, a lot of research on measuring object distances was carried out in developing security systems and surveillance systems, one of which was in security in the environment of many human objects or crowds. This study uses the object segmentation method using the Histogram of Oriented Gradient feature to segment crowd objects. In determining the value of the distance based on information using a segmented object centroid. Calculations are performed using the Euclidian Distance calculation method to find the shortest distance between the centroid of the bounding box and the camera. The results of this study from object distance can distinguish human objects that have crowds with the best accuracy with a measurement error of 5.7%. The research have conclusion that the main findings produced can be used to produce an accurate human crowd object recognition system that is able to provide information on the value of the object's distance to the camera when the object.