TY - GEN
T1 - Explicit image detection using YCbCr space color model as skin detection
AU - Basilio, Jorge Alberto Marcial
AU - Torres, Gualberto Aguilar
AU - Pérez, Gabriel Sánchez
AU - Medina, L. Karina Toscano
AU - Meana, Héctor M.Pérez
PY - 2011
Y1 - 2011
N2 - In this paper a novel way to detect explicit content images is proposed using the YCbCr space color, moreover the color pixels percentage that are within the image is calculated which are susceptible to be a tone skin. The main goal to use this method is to apply to forensic analysis or pornographic images detection on storage devices such as hard disk, USB memories, etc. The results obtained using the proposed method are compared with Paraben's Porn Detection Stick software which is one of the most commercial devices used for detecting pornographic images. The proposed algorithm achieved identify up to 88.8% of the explicit content images, and 5% of false positives, the Paraben's Porn Detection Stick software achieved 89.7% of effectiveness for the same set of images but with a 6.8% of false positives. In both cases were used a set of 1000 images, 550 natural images, and 450 with highly explicit content. Finally the proposed algorithm effectiveness shows that the methodology applied to explicit image detection was successfully proved vs Paraben's Porn Detection Stick software.
AB - In this paper a novel way to detect explicit content images is proposed using the YCbCr space color, moreover the color pixels percentage that are within the image is calculated which are susceptible to be a tone skin. The main goal to use this method is to apply to forensic analysis or pornographic images detection on storage devices such as hard disk, USB memories, etc. The results obtained using the proposed method are compared with Paraben's Porn Detection Stick software which is one of the most commercial devices used for detecting pornographic images. The proposed algorithm achieved identify up to 88.8% of the explicit content images, and 5% of false positives, the Paraben's Porn Detection Stick software achieved 89.7% of effectiveness for the same set of images but with a 6.8% of false positives. In both cases were used a set of 1000 images, 550 natural images, and 450 with highly explicit content. Finally the proposed algorithm effectiveness shows that the methodology applied to explicit image detection was successfully proved vs Paraben's Porn Detection Stick software.
KW - Explicit content
KW - Pattern recognition
KW - Skin detection
KW - YCbCr color space
UR - http://www.scopus.com/inward/record.url?scp=79958723400&partnerID=8YFLogxK
M3 - Contribución a la conferencia
AN - SCOPUS:79958723400
SN - 9789604742707
T3 - Applications of Mathematics and Computer Engineering - American Conference on Applied Mathematics, AMERICAN-MATH'11, 5th WSEAS International Conference on Computer Engineering and Applications, CEA'11
SP - 123
EP - 128
BT - Applications of Mathematics and Computer Engineering - American Conference on Applied Mathematics, AMERICAN-MATH'11, 5th WSEAS International Conference on Computer Engineering and Applications, CEA'11
T2 - American Conference on Applied Mathematics, AMERICAN-MATH'11, 5th WSEAS International Conference on Computer Engineering and Applications, CEA'11
Y2 - 29 January 2011 through 31 January 2011
ER -