Segmentasi Video Berbasis Objek Bergerak Dengan Improved Three Frame Difference Algorithm

Charisma, Praditya Lutvi (2020) Segmentasi Video Berbasis Objek Bergerak Dengan Improved Three Frame Difference Algorithm. Undergraduate thesis, Institut Teknologi Sepuluh Nopember.

[img] Text
06111640000055-Undergraduate_Thesis.pdf - Accepted Version
Restricted to Repository staff only

Download (2MB) | Request a copy

Abstract

Rekaman CCTV dimanfaatkan untuk pengecekan peristiwa di area pengawasan. Namun, proses pelacakan membutuhkan waktu lama karena harus mengamati video dari awal, dan mengamati video tanpa aktifitas objek pada interval waktu tertentu. Oleh karena itu, pada penelitian ini dilakukan pemotongan video yang hanya berisi aktifitas objek. Pada penelitian ini, digunakan improved three frame difference algorithm untuk didapatkan interval waktu terjadinya gerakan objek pada video dengan melakukan deteksi objek bergerak. Selanjutnya, dilakukan segmentasi video untuk memudahkan pemilik video dalam melakukan pengamatan atau pengecekan rekaman cctv. Pada penelitian ini, dilakukan uji coba terhadap 4 video dengan spesifikasi 4 parameter antara lain: selang frame sebesar 10, threshold sebesar 30, size strel sebesar 10x10, dan luas minimum sebesar 100. Dari hasil uji coba, didapatkan nilai f1 score terendah sebesar 82,9% dan perhitungan accuracy terendah sebesar 97,8%. Rata-rata, nilai f1 score segmentasi naik sebesar 21,7% daripada hanya menggunakan algoritma three frame difference. ================================================================================================ CCTV recordings are used to check activities which occur in an area of surveillance. However, the tracking process takes time since we need to play the video from the beginning, and we may check on unimportant at certain time intervals. Therefore, this study aims to trim the video that only contain the moving object. This study employes the improved three frame difference algorithm to obtain the time intervals of moving object on the video by detecting its movement. Furthermore, we performance video segmentation to facilitate the authorities either to observe or to check cctv recordings. In this study, we experiment with 4 videos, by specifying 4 parametes : frame intervals is 10, threshold is 30, size strel is 10x10, and minimum area is 100. In the experimental result we obtain the lowest value of F1 Score which is 82.9% and the lowest value of accuracy which is 97,8%. In average, the f1 score value of segmentation increases 21,7% than only three frame difference algorithm.

Item Type: Thesis (Undergraduate)
Uncontrolled Keywords: segmentasi video, deteksi objek bergerak, improved three frame difference algorithm
Subjects: Q Science > QA Mathematics > QA76.6 Computer programming.
Q Science > QA Mathematics > QA76.76.A65 Application software. Enterprise application integration (Computer systems)
T Technology > TA Engineering (General). Civil engineering (General) > TA1637 Image processing--Digital techniques
Divisions: Faculty of Mathematics, Computation, and Data Science > Mathematics > 44201-(S1) Undergraduate Thesis
Depositing User: Praditya Lutvi Charisma
Date Deposited: 26 Aug 2020 04:08
Last Modified: 26 Aug 2020 04:08
URI: http://repository.its.ac.id/id/eprint/79702

Actions (login required)

View Item View Item