Sistem Pengontrol Presentasi Menggunakan Pengenalan Gestur Tangan Berbasis Fitur pada Contour dengan Metode Klasifikasi Support Vector Machine

Muhammad Hafid Khoirul (Fakultas Ilmu Komputer, Universitas Brawijaya)
Fitri Utaminingrum (Fakultas Ilmu Komputer, Universitas Brawijaya)

Jurnal Pengembangan Teknologi Informasi dan Ilmu Komputer
Vol 4 No 4 (2020): April 2020

In recent years, presentation has been a part of our lives. Presenter is limited by mouse, keyboard, or any other devices to control presentation. By using those devices conclude that less effective, less efficient, and less natural. On another hand, further method to control presentation uses hand gesture. It improves thinking process since the method can facilitate to give visual guide and deliver the presentation more effective, more efficient, and more natural. The developed system can recognize hand gestures and give output to control presentation based on the recognized hand gesture. Contour features such as Hu moments, circularity,convexity, aspect ratio, ratio of contour area, ratio of centroid height and width, and white pixel ratio used to distinguish every hand gestures. HSV and YCbCr color spaces used for skin segmentation. To recognized hand gestures used Support Vector Machine classification method. As a result, by using contour features and SVM to recognize hand gestures, system obtained average accuracy time of 88.15% and average computation time of 0.1567s.