<![CDATA[Facial recognition research is usually done with a rectangular frame model. previously it is not known whether this model is the best model among other models in face recognition. This research conducted a comparison of face recognition accuracy with elliptical frame model, triangle, rectangle, and pentagon. RGB image is used for facial segmentation, and skin color is used as a feature. using 320 x 240 pixel image, face position is front view, shooting distance about 200 cm, and use one color for background. The method of recognition used is Template Matching. Test results on 40 test images is the highest recognition accuracy level is a rectangular frame that is 97%, and the lowest accuracy is the triangle frame that is equal to 68%.Keywords : Face recognition, Frame, Segmentation, Template Matching]]>
Copyrights © 0000
