Lunar Crater Detection & Recognition

Period: September 2014 - June 2015

The Lunar Crater Detection & Recognition project was an undergraduate senior design software project developed by a team of computer science students of the College of Engineering, Computer Science and Technology (ECST) in California State University of Los Angeles (Cal State LA) and sponsored by Jet Propulsion Laboratory (JPL). The developers were computer science undergraduates, Natalie Gallegos, Tony Liang, Marvin Mendez, Raul Ornelas, and Alberto Serrano Jr. Their supervisor was Professor Elaine Kang. Their advisors were graduate students, Richard Cross and Saman Saeedi. Emily Law acted as a liaison between JPL and the developers, their supervisor, and their advisors.

For humans and robots to travel to the Moon, safe spacecraft landings on the Moon is essential. It requires carefully mapping out the best lunar terrain. However, there are many craters on the Moon. The detection and avoidance of the craters are critical for landing on the Moon.

The Lunar Mapping and Modeling Portal (LMMP) team at JPL entrusted the student developers to develop an algorithm that automatically detects lunar craters in images provided by the Lunar Reconnaissance Orbiter Camera (LROC). They wanted to prioritize the algorithm’s accuracy over performance. So, the students’ approach was to combine several algorithms to achieve high accuracy results. For crater detection and recognition, Raul and Albert implemented the ellipse fitting algorithm, Marvin implemented the circle Hough transform algorithm, and Saman implemented template matching and a machine learning algorithm. Natalie and Tony implemented an algorithm that computes a crater’s diameter and depth. The developers combined the algorithms and integrated them to develop the software called Ringtoss.