An Application: Vision-Guided Navigation Using IGA

Abstract

In general, the purpose of a depth recovery algorithm is to determine the locations of objects in the field of view. However, for some tasks, such as navigation, it is equally important to determine where objects aren’t located (i.e. navigable or free space). This information is inherently included in the depth estimates provided by the depth recovery algorithm: if an object at image location (i, j) is identified as lying a distance d) from the observer, it is implicit that there are no objects at image location (i, j) which lie less than distance d) to the observer.