SENSOR FUSION USING DETECTOR CONFIDENCE BOOSTING

摘要

A method for detecting one or more target objects is provided including obtaining 2-dimensional imaging information and 3-dimensional point cloud information of a target zone. The method also includes determining a ground plane in the point cloud information and removing the ground plane to generate modified 3-dimensional information. Also, the method includes identifying a set of 2-dimensional candidate objects from the 2-dimensional imaging information, and identifying a set of 3-dimensional candidate objects from the modified 3-dimensional information. The method also includes determining, for each of at least some of the 2-dimensional candidate objects, a corresponding 3-dimensional candidate object from the set of 3-dimensional candidate objects. Further, the method includes modifying the 2-dimensional confidence measure for each of the at least some of the 2-dimensional candidate objects to generate fused confidence measures based on whether the 2-dimensional candidate object corresponds to a 3-dimensional candidate object.

主权项

1. A method for detecting one or more target objects, the method comprising:
obtaining 2-dimensional imaging information of a target zone; obtaining 3-dimensional point cloud information of the target zone; determining a ground plane in the 3-dimensional point cloud information and removing the ground plane to generate modified 3-dimensional information; identifying a set of 2-dimensional candidate objects from the 2-dimensional imaging information, each 2-dimensional candidate object having associated therewith a corresponding 2-dimensional confidence measure; identifying a set of 3-dimensional candidate objects from the modified 3-dimensional information by clustering proximal points from the modified 3-dimensional information into object groups, each 3-dimensional candidate object having associated therewith a corresponding 3-dimensional confidence measure; determining, for each of at least some of the 2-dimensional candidate objects, a corresponding 3-dimensional candidate object from the set of 3-dimensional candidate objects; and modifying the 2-dimensional confidence measure for each of the at least some of the 2-dimensional candidate objects based on whether the 2-dimensional candidate object corresponds to a 3-dimensional candidate object, to generate fused confidence measures using the 2-dimensional confidence measures for each of the at least some of the 2-dimensional candidate objects and the 3-dimensional confidence measures of the determined corresponding 3-dimensional candidate objects, whereby the fused confidence measure filters out false objects from the 2-dimensional candidate objects to identify with a high degree of confidence a set of target objects.