JACIII Vol.7 No.3 pp. 289-293
doi: 10.20965/jaciii.2003.p0289


Real Time Adaptive Color Segmentation for Mars Landing Site Identification

Tuan A. Duong and Vu A. Duong

Bio-Inspired Technologies and Systems Group, Avionic Systems and Technology Division
Jet Propulsion Laboratory, California Institute of Technology, Pasadena, CA 91109 U.S.A.

September 10, 2003
August 26, 2003
October 20, 2003
color segmentation, adaptive learning, landing site identification, adaptive architecture, real time identification, safe landing, cascade error projection, limited weight quantization
Real time identification of planetary landing sites is of significant importance in NASA’s precision landing program. As a lander descends towards a potential landing site, it is important to determine whether or not the potential site is free of debris to allow safe landing. This requires real time processing of images acquired by the lander during the approach, so that appropriate navigational corrections can be made to direct the lander to a safe landing zone. In this paper we discuss an adaptive color segmentation technique that can aid in identifying safe landing terrain, and terrain that may be rock covered, dusty, and unfavorable to land. A new learning architecture that allows real time adaptation in a dynamically changing environment as the lander approaches a landing site is evaluated. Results indicate that a real time adaptive color segmentation approach is sufficient to identify safe landing zones. The paper also discusses the time required for adaptation, a critical parameter during an actual descent. The simulation-based case study reported in this paper is a primary step toward developing a more realistic case for landing site identification.
Cite this article as:
T. Duong and V. Duong, “Real Time Adaptive Color Segmentation for Mars Landing Site Identification,” J. Adv. Comput. Intell. Intell. Inform., Vol.7 No.3, pp. 289-293, 2003.
Data files:

*This site is desgined based on HTML5 and CSS3 for modern browsers, e.g. Chrome, Firefox, Safari, Edge, Opera.

Last updated on Jun. 03, 2024