An Algorithm for the Identification and Measurement of Jagged Appertures from a Noisy Image

In this paper a combinatorial algorithm is presented and implemented for the identification and measurement of apertures from a noisy image. These apertures appear in the form of highly irregular shapes, and are approximated by circular disks. Three circular disks are defined for each aperture. The indisk which the largest disk contained in the aperture, the outdisk which is the smallest disk that contains the aperture and with center that of the indisk, and the approximation disk with center that of the indisk and area equal to the area of the aperture. The problem of identifying jagged apertures from a noisy image that could be approximated by circular disks has many applications in diverse areas. In this paper we will present how this problem emerges in meteorology, where hail pads have to be analyzed in order to determine the number and characteristics of the hail stones that collided with the pad. Computational results on a set of benchmark images from actual data are presented