Title :
Ensemble Threshold Segmentation for hand detection
Author :
Rungruangbaiyok, Sunthorn ; Duangsoithong, Rakkrit ; Chetpattananondh, Kanadit
Author_Institution :
Dept. of Electr. Eng., Prince of Songkla Univ., Songkhla, Thailand
Abstract :
Hand detection can be used for human computer interaction (HCI) and has many methods that use to detect the hand image. Skin color detection is one of the methods however there are many types of color model with threshold optimization problem. This paper proposes an Ensemble Threshold Segmentation method (ETS) to improve the segmentation technique. The ETS was implemented to optimize threshold of any color model to compare the accuracy with existing color models and well-known Otsu´s method. The results show that the best accuracy technique is the ETS with HSV color model.
Keywords :
human computer interaction; image colour analysis; image segmentation; optimisation; ETS; HCI; HSV color model; ensemble threshold segmentation; ensemble threshold segmentation method; hand detection; hand image detection; human computer interaction; skin color detection; threshold optimization problem; Accuracy; Colored noise; Gesture recognition; Image color analysis; Image segmentation; Mathematical model; Skin; hand detection; hand image segmentation; skin color; thresholding;
Conference_Titel :
Electrical Engineering/Electronics, Computer, Telecommunications and Information Technology (ECTI-CON), 2015 12th International Conference on
Conference_Location :
Hua Hin
DOI :
10.1109/ECTICon.2015.7207071