Pattern-Recognition System for Approaching a Known Target

NASA’s Jet Propulsion Laboratory, Pasadena, California

Saturday, March 01 2008 During the third phase, distinctive pattern of six stripes on the ramps is utilized in a close-range algorithm for positioning the rover within 10 cm of the bottom of the ramps. At the beginning of this phase, when the rover is ≈5 m from the lander, the rover circles around the lander. The close-range algorithm includes three major steps: feature extraction, feature match, and pose estimation. The features are the six stripes, which are arranged so that any two of them is a topologically and spatially unique combination, so as to greatly reduce uncertainty and computational complexity.

An edge-detection subalgorithm is applied first. Then all straight-line segments are extracted. In order to find the stripes, the close-range algorithm looks for the ramps, which are defined by a set of long straight and nearly parallel lines (see figure). When a single stripe is detected in the image, a linear affine transformation based on its four corners can be constructed. If transformation provides a correct match with a known stripe, it can help to find other matches. A search on all stripes is performed to find the best matches. Once the matches are found, the position and orientation of the rover relative to the lander are estimated by use of the outside corners of the stripes. A minimum of four stripes is used to ensure safe navigation.

This work was done by Terrance Huntsberger and Yang Cheng of Caltech for NASA’s Jet Propulsion Laboratory.

The software used in this innovation is available for commercial licensing. Please contact Karina Edmonds of the California Institute of Technology at (626) 395-2322. Refer to NPO-41867.