Three dimensional object recognition
Object recognition is a complex problem in computer vision. In most recognition systems, features are extracted from sensors such as intensity, tactile, and range. These features are matched to a database of modeled objects in an attempt to determine which object(s) are present. Once the object identities are known, the orientation of each object relative to some base frame of reference is determined. A solution for recognizing polyhedral objects using surface normals as the sole input feature is given. This technique exploits strong constraints on the angles between the faces of an object to perform recognition. Exhaustive experiments involving the of use of all possible combinations of input to the system have yielded encouraging results. The system mentioned uses only surface normals which could be derived from a single sensor. However, multiple sensors can be used to make recognition easier by providing different features. The addition of multiple sensors introduces the problem of combining features which are sometimes contradictory. A Bayesian technique for fusing data between sensors, of any type is proposed. This method adjusts the confidence we have in each feature based on the support from multiple sensors. The confidence values aid in the elimination of noisy input to a recognition system making those systems more robust.
This item is only available in print in the UCF Libraries. If this is your thesis or dissertation, you can help us make it available online for use by researchers around the world by downloading and filling out the Internet Distribution Consent Agreement. You may also contact the project coordinator Kerri Bottorff for more information.
Bachelor of Science (B.S.)
College of Arts and Sciences
Arts and Sciences -- Dissertations, Academic;Dissertations, Academic -- Arts and Sciences
Length of Campus-only Access
Honors in the Major Thesis
Lavoie, Matt J., "Three dimensional object recognition" (1991). HIM 1990-2015. 2.