Feature space trajectory methods for active object recognition

We advance new active object recognition algorithms that classify rigid objects and estimate their pose from intensity images. Our algorithms automatically re-position the sensor if the class or pose of an object is ambiguous in a given image (our algorithms automatically detect ambiguity) and incorporate data from multiple object views in determining the final object classification and pose estimate. A probabilistic feature space trajectory (FST) in a global eigenfeature space is used to represent 3-D distorted views of an object and to estimate the class and pose of an input object. Assuming that an observed feature vector consists of Gaussian noise added to a point on the FST, we derive a probability density function (pdf) for an observation conditioned on the class and pose of the object; we use Bayesian estimation and hypothesis testing theory to show that FST estimates are approximations to the maximum a posteriori probability pose estimate and the minimum probability of error classifier. Confidence measures for the class and pose estimates, derived using the probabilistic FST object representation, determine when additional observations are required as well as where the sensor should be positioned to provide the most useful information. We present various uses of our active object recognition algorithm in automated assembly and inspection problems. We demonstrate the ability to use FSTs constructed from images rendered from CAD models to recognize real objects in real images. We present test results for a set of metal machined parts, with specular surfaces, and for an assembly problem consisting of several parts. We also show the ability of the FST to learn and add new objects.