Edge Detection using MINMAX Measures

S. Ezekiel and M. Lang (USA)


Edges, fractal dimension, minmax measure, slope image


In this paper, we present a minmax measure-based algorithm for edge detection for images. Edge detection is a tool that has been widely used in image processing and computer vision for a variety of reasons. In an image, edges are areas where there is a strong change in contrast. Typically, edges represent the boundaries of objects within an image, and therefore, determining the locations of these boundaries is important in further machine analysis of image content. In most cases, edge detection requires smoothing and differentiation of the image. Smoothing results in the loss of image information, and differentiation is an ill-conditioned problem. In this paper, we calculate the local fractal dimension to estimate the roughness of that region. To calculate the local fractal dimension, we use the minmax measure. We then form an image, called the slope image, from the fractal dimension. We then apply simple threshold techniques on the slope image to extract edge information. The results suggest that this method is more effective than traditional methods and that is has the capability to be applied to a broad range of image categories.

Important Links:

Go Back