Paper- Restricted to Campus Access
Physics & Astronomy
The problem we are seeking to solve with this research is how we can develop an efficient and effective computer program to automatically count impact craters. The rationale behind this is to automate and increase the speed for the process by which scientists can count craters on the surface of a planetoid for surface age-dating. This time-intensive, hand-counting process can involve upwards of tens of thousands of craters, particularly on older surfaces like the Moon or Mars (e.g. Baker 2001, Bierhaus, et al. 2005, Martin-Wells, et al. 2017, McEwen, et al. 2005, Robbins, et al. 2011). Several known challenges must be overcome in order to create such a system. Background features and uneven, often deformed, edges of the craters make this task difficult. Variable lighting conditions also present difficulties for programs analyzing such surfaces. Due to energy scaling, cratered surfaces are naturally fractal, a property our method uses to its advantage (Melosh 1989, Vickery 1986). Over the course of the Summer Fellows program, we have begun to write code in the Interactive Data Language (IDL) to process large image datasets. To date, we are in the process of defining the smoothing algorithm that will be used as the first part of the automatic counting process. As we continue testing, the program will be designed to utilize the fractal nature of the images, applying a scale-independent convolution filter to the smoothed images at each fractal scale.
Cloward, Micah, "Machine Counting of Impact Cratering" (2018). Physics and Astronomy Summer Fellows. 20.