Data Mining System for Planetary Images - Crater Detection and Categorization
Rie Honda, Osamu Konishi, Ryushi Azuma, H. Yokogawa, S. Yamanaka, Yuta Iijima · 2000
The surfaces of solid planets are covered with a tremendous number of craters of various shape and size that reflect the geological history and rheology of the planetary surfaces. In the present study we examined crater detection and categorization system for planetary images in view of data mining from a large-scale scientific image database. Crater detection by combinational Hough transforms and generic algorithm and categorization by self-organizing mapping are investigated as component processes in detail. 1. Introduction A large amount of remote-sensing image data has recently been obtained by satellites due to the improvement of sensors and telemetry systems. These conditions have led scientists to request the construction of a large-scale scientific image database, to extract the characteristics of these images, and to construct a secondary database without manual operation. Studies aimed at the automatic extraction of objects of interest and classification based on...