Posture recognition in ballet dance a case study on fuzzy uniform discrete membership function

The proposed work gives an overview of fuzzy discrete membership function based algorithm that would automatically recognize an unknown ballet posture from seventeen Ballet dance primitives. At first, skin color of the dancer is extracted from the RGB input image. Then skeleton of the given postures is generated after performing morphological skeleton and spur operation on them. Straight line approximation and polygon represented code are produced in the later stages. In the last stage, a fuzzy discrete membership function based on code is measured and with the help of similarity function, unknown posture is recognized.