Monitoring Keramaian Menggunakan Crowd Counting CNN

Khairani, Puti Syifa (2022) Monitoring Keramaian Menggunakan Crowd Counting CNN. Other thesis, Institut Teknologi Sepuluh Nopember.

Text 07211840000077-Undergraduate_Thesis.pdf Restricted to Repository staff only Download (8MB)

Abstract

Di era teknologi ini, Artificial Intelligence (AI) atau kecerdasan buatan sudah menjadi bagian dari kehidupan sehari-hari. Artificial Intelligence menghasilkan sistem atau teknologi yang cenderung bekerja lebih cepat, akurat, dan minim kesalahan (human error). Belajar dari pendekatan yang dilakukan sebelumnya selama puncak pandemi COVID-19, pemantauan terhadap kepatuhan masyarakat terhadap protokol kesehatan, terutama terkait menghindari kerumunan, masih didapatkan dari laporan petugas di lapangan. Pendekatan ini kurang efektif. Dalam penelitian tugas akhir ini dikembangkan metode untuk menghitung jumlah orang dalam suatu area keramaian hasil tangkapan kamera menggunakan Convolutional Neural Network (CNN) yaitu CSRNet. Performa model cukup baik dengan MAE yang dihasilkan 40% dan 60% (model A dan B) lebih rendah dibandingkan dengan beberapa metode penelitian sebelumnya. Model ini juga dievaluasi dengan data video dengan menghitung jumlah frame yang dapat diproses dalam satu detik (FPS).
==============================================================================================================================
In this technological era, Artificial Intelligence (AI) has become a part of everyday life. Artificial Intelligence produces a system or technology that tends to work faster, more accurately, and with minimal human errors. Learning from the approach taken during the peak of the COVID-19 pandemic, monitoring community compliance with health protocols, especially regarding the avoidance of crowds, was analyzed through reports from officers on the spot. This approach is less effective and might be inaccurate. In this undergraduate final project, a method was developed to calculate the number of people in a crowded area captured by the camera using the Convolutional Neural Network (CNN), namely CSRNet. Results show that the performance of the model is quite good, with the MAE produced being 40% and 60% (model A and B) lower than several previous research methods. This model is also evaluated with video as data input by calculating the number of frames that can be processed per second (FPS).

Item Type:	Thesis (Other)
Additional Information:	RSKom 006.32 Kha m-1 2022
Uncontrolled Keywords:	Crowd Monitoring, Crowd Counting, Convolutional Neural Network. Crowd Monitoring, Crowd Counting, Convolutional Neural Network.
Subjects:	Q Science > QA Mathematics > QA75 Electronic computers. Computer science. EDP
Divisions:	Faculty of Intelligent Electrical and Informatics Technology (ELECTICS) > Computer Engineering > 90243-(S1) Undergraduate Thesis
Depositing User:	Mr. Marsudiyana -
Date Deposited:	17 Jun 2026 07:12
Last Modified:	17 Jun 2026 07:12
URI:	http://repository.its.ac.id/id/eprint/133860

Actions (login required)

View Item