Perbandingan Metode Hog dengan Haar Like Feature pada Implementasi Human Traking pada Kamera Cctv

Prastio, Dedi (2020) Perbandingan Metode Hog dengan Haar Like Feature pada Implementasi Human Traking pada Kamera Cctv. Undergraduate thesis, Universitas Internasional Batam.

[img]
Preview
Text
s-1621001-abstract-en.pdf

Download (844kB) | Preview
[img]
Preview
Text
s-1621001-abstract-id.pdf

Download (844kB) | Preview
[img]
Preview
Text
s-1621001-bibliography.pdf

Download (1MB) | Preview
[img]
Preview
Text
s-1621001-chapter1.pdf

Download (1MB) | Preview
[img]
Preview
Text
s-1621001-chapter2.pdf

Download (2MB) | Preview
[img] Text
s-1621001-chapter3.pdf
Restricted to Repository staff only

Download (2MB) | Request a copy
[img] Text
s-1621001-chapter4.pdf
Restricted to Repository staff only

Download (2MB) | Request a copy
[img]
Preview
Text
s-1621001-chapter5.pdf

Download (843kB) | Preview
[img]
Preview
Text
s-1621001-cover-id.pdf

Download (44kB) | Preview

Abstract

Saat ini perangkat yang digunakan sebagai keamanan rumah sangat beragam diantaranya adalah penggunaan atau pemasangan Closed Circuit Television (CCTV) di posisi tertentu di rumah. Saat ini kamera CCTV yang banyak digunakan adalah Fix CCTV, dinama pada kamera CCTV ini memiliki kelemahan adanya Blind Spot (titik buta), sehingga pada titik tersebut kamera tidak dapat merekam atau mendeteksi adanya orang. Oleh karena itu pada penelitian ini dirancang prototipe kamera CCTV yang mampu bergerak sesuai dengan tracking (pendeteksian) manusia dengan Metode Histogram of Oriented Gradients (HOG) dan Haar Like Feature. Pada penelitian ini akan membandingkan hasil tracking manusia dengan metode HOG dan metode Haar Like serta keakuratan respon dari kamera CCTV terhadap deteksi manusia tersebut. Dalam perancangan prototipe terdiri dari, Raspberry Pi 3 Model B, kamera CCTV seri THC-T120-P dan motor servo. Kamera CCTV sebagai sensor utama yang ditambahkan dengan motor servo sebagai penggeraknya, kamera akan mendeteksi object yang selanjutnya diproses dengan metode HOG atau Haar like sebagai pendeteksi manusia di dalam Raspberry Pi 3, jika terdeteksi manusia maka kamera CCTV akan bergerak mengikuti pergerakan orang (perpindahan posisi dari orang) tersebut. Hasil pengujian yang dilakukan menunjukan bahwa deteksi manusia dengan menggunakan metode HOG menghasilkan persentase pendeteksian sebesar 57.14% sedangkan menggunakan metode Haar like menghasilkan persentase pendeteksian sebesar 71.43%, kondisi ini dikarenakan pengaruhnya posisi object. Namun dalam control pergerakan kamera CCTV telah berhasil dimana telah dapat mengikuti object yang terdeteksi. ********************************************************************** Currently, the devices used as home security are very diverse including the use or installation of Closed Circuit Television (CCTV) in certain positions at home. Presently, CCTV cameras widely used are Fix CCTV, which has the weakness of the Blind Spot, consequently at that point, the camera cannot record or detect the presence of people. Therefore, in this study, a prototype CCTV camera was designed to be able to move according to human tracking (Histogram of Oriented Gradients) (HOG) and Haar Like Feature methods. This research compared the results of human tracking with the HOG method and the Haar Like method as well as the accuracy of the response from the CCTV camera to the human detection. The design of the prototype consisted of Raspberry Pi 3 Model B, CCTV camera series THC-T120-P and servo motor. CCTV cameras as the main sensor were added with servo motor as the driving force. The camera detected an object which is then processed by the HOG or Haar like method as a human detector in Raspberry Pi 3. If detected by humans then the CCTV camera would shove to follow the movement of people (movement of the position of the person). The results of tests conducted show that human detection using the HOG method produces a detection percentage of 57.14% while using the Haar like method produces a detection percentage of 71.43%, this condition is due to the effect of the object's position. But in the control of the movement of CCTV cameras have been successful where it can follow the detected object.

Item Type: Thesis (Undergraduate)
Additional Information: Similarity: 18
Uncontrolled Keywords: CCTV Cameras, Raspberry Pi, HOG Method, Haar Like Method, Human Detection
Subjects: T Technology
T Technology > Automation
Divisions: School of Industrial Technology > Electrical Engineering
SWORD Depositor: Admin Repository Universitas Internasional Batam
Depositing User: Admin Repository Universitas Internasional Batam
Date Deposited: 28 Apr 2020 09:19
Last Modified: 28 Apr 2020 09:24
URI: http://repository.uib.ac.id/id/eprint/2406

Actions (login required)

View Item View Item