Al Zulmi, Kahlil Gibran (2026) Perancangan Perangkat Lunak untuk Ekstraksi Data Titik Pandang dari Rekaman Layar Pelacakan Gerak Mata Berbasis Visi Komputer. Other thesis, Institut Teknologi Sepuluh Nopember.
![[thumbnail of 5049221015-Undergraduate_Thesis.pdf]](http://repository.its.ac.id/style/images/fileicons/text.png) |
Text
5049221015-Undergraduate_Thesis.pdf - Accepted Version Restricted to Repository staff only Download (3MB) | Request a copy |
Abstract
Pelacak gerak mata kelas konsumen, seperti Tobii Eye Tracker 5 yang digunakan untuk pengalaman bermain gim yang lebih imersif, menawarkan aksesibilitas tinggi untuk penelitian di bidang Teknologi Kedokteran. Namun, perangkat lunak bawaannya, Tobii Experience dan Tobii Ghost, tidak menyediakan dukungan untuk mengekspor data mentah koordinat titik pandang (gaze point). Perancangan ini bertujuan mengembangkan perangkat lunak berbasis Python dan OpenCV untuk mengekstraksi data koordinat (x, y) dari rekaman layar lapisan visual (overlay) pelacak mata tersebut. Metodologi melibatkan pengolahan video rekaman menggunakan beberapa pendekatan algoritma deteksi: berbasis warna, kontur, blob, serta Hough Circle Transform (HCT) yang dikombinasikan dengan Kalman Filter. Validasi dilakukan dengan membandingkan hasil deteksi terhadap video stimulus terprogram (ground truth) untuk mengukur tingkat keberhasilan deteksi dan galat posisi (Euclidean distance), diikuti dengan evaluasi menggunakan gim interaktif berbasis matematika sederhana. Hasil pengujian menunjukkan bahwa metode HCT dan HCT dengan Kalman Filter mencapai performa terbaik dengan tingkat keberhasilan deteksi 100%, secara signifikan unggul dibandingkan metode berbasis warna yang hanya mencapai 8%. Rata-rata galat Euclidean distance yang dihasilkan adalah 84,8 piksel. Meskipun belum memenuhi standar presisi tinggi untuk diagnosis medis klinis, perangkat lunak ini menunjukkan viabilitas fungsional sebagai alat bantu penelitian perilaku visual yang terjangkau untuk analisis perhatian kasar (coarse-grained) dan interaksi manusia-komputer sederhana.
=================================================================================================================================
Consumer-grade eye trackers, such as the Tobii Eye Tracker 5 used for playing games more immersively, offer high accessibility for research in the Medical Technology field. However, their built-in software, Tobii Experience and Tobii Ghost, do not provide support for exporting raw gaze point coordinates. This design aims to develop Python and OpenCV-based software to extract coordinate data (x, y) from screen recordings of the eye tracker's visual overlay. The methodology involves processing video recordings using several detection algorithms: color-based, contour-based, blob detection, and Hough Circle Transform (HCT) combined with a Kalman Filter. Validation was performed by comparing detection results against a programmed stimulus video (ground truth) to measure detection success rates and positional error (Euclidean distance), followed by an evaluation using a simple math-based interactive game. The results showed that the HCT and HCT with Kalman Filter methods achieved the best performance with a 100% detection rate, significantly outperforming the color-based method, which only reached 8%. The average Euclidean distance error was 84.8 pixels. Although it does not yet meet high-precision standards for clinical medical diagnosis, this software demonstrates functional viability as an affordable tool for visual behavior research focusing on coarse-grained attention analysis and simple human-computer interaction.
| Item Type: | Thesis (Other) |
|---|---|
| Uncontrolled Keywords: | pelacak gerak mata, visi komputer, ekstraksi data, OpenCV, Tobii Eye Tracker 5, eye tracker, computer vision, data extraction, OpenCV, Tobii Eye Tracker 5 |
| Subjects: | T Technology > T Technology (General) > T57.5 Data Processing T Technology > TA Engineering (General). Civil engineering (General) > TA1637 Image processing--Digital techniques. Image analysis--Data processing. |
| Divisions: | Faculty of medicine and health (MEDICS) > Medical Technology > 11503-(S1) Undergraduate Thesis |
| Depositing User: | Kahlil Gibran Al Zulmi |
| Date Deposited: | 03 Feb 2026 01:43 |
| Last Modified: | 03 Feb 2026 01:43 |
| URI: | http://repository.its.ac.id/id/eprint/131693 |
Actions (login required)
 |
View Item |