Multiple layered Hierarchical Feature Tracking for Grove Scene
SangHoon Han, Thomas Francis Burks · 2010 Pittsburgh, Pennsylvania, June 20 - June 23, 2010 · 2010
This study presents a robust feature tracking method specified for grove scenes. Unlike artificial structure scenes which have salient features or surfaces, it is not easy to apply ordinary feature detection methods to natural objects such as those found in a grove scene. Even though each extracted feature has a different degree of intensity from other features, a single image processing equally deals with all of the extracted features. It is vulnerable to disturbances caused by flickering in the feature detection. It is necessary to take stronger features into account prior to weaker features.