Penerapan Algoritma Haar Like Feature Untuk Deteksi Pornografi

Abstract

In this study, researchers tried to apply the digital image method, namely the Haar-Like Feature algorithm, hm, to detect objects in pornographic videos. The input space given to the Haar-Like Feature is in the form of the x-axis, y-axis, area width, and area height that has been processed through the image labeling stage; the input space will be used to detect and determine the position of the object. The objects seen in this study are breasts and sex toys, based on a survey of 102 questions conducted by researchers who got 70 respondents. The survey results said the highest percentage of porn on the breast is 60% and sex toys 8%. The conclusion from this research is that the use of the haar-like feature algorithm can detect objects in pornographic videos and the results of the evaluation of the model with a training dataset of 4881 images get an accuracy value of 63.33%.