This paper presents an efficient method for extracting the skeleton of a planar shape. The method is based on computing local symmetries of the shape’s linearized contour. The centrelines of the local symmetries form the skeleton of the shape. The method is fast when compared to existing skeletonization methods. The speed-up is usually a factor of more than 10 times for high-resolution images. The method is also robust against noise and geometrical transformations, such as rotation and uniform scaling.