FILTER HALAMAN WEB PORNOGRAFI MENGGUNAKAN KECOCOKAN KATA DAN DETEKSI WARNA KULIT
ABSTRACT: This paper presents
an effort to detect pornographic webpages. It was stated that a positive
relationship exists between percentage of human skin color in an image and the
image itself (Jones et.al., 1998). Based on the statement, rather than using
the traditional method of text-filtering, this paper propose a new approach to
detect pornographic images by using skin color detection. The skin color
detection performed by using RGB, HSI, and YCbCr color model. Using algorithm
stated by Ap-apid (Ap-apid, 2005), the system will classify nude and not-nude
images. If one or more nude images are found, the system will block the
webpage. Keywords: Webpage Filtering, Image Processing, Pornography, Nudity,
Skin Color, Nude Images
Penulis: Yusron Rijal, Awalia
Nofitasari
Kode Jurnal: jpmatematikadd110138
