Browsing by Author "Dr. Wesley E. Snyder, Committee Chair"

Now showing 1 - 3 of 3

An Accumulative Framework for Object Recognition
(2009-04-21) Krish, Karthik; Dr. Wesley E. Snyder, Committee Chair; Dr. Hamid Krim, Committee Member; Dr. Griff Bilbro, Committee Member; Dr. Siamak Khorram, Committee Member; Dr. Benjamin Watson, Committee Member
Object recognition has received a lot of attention over the years and has reached a level where we have a lot of algorithms which can identify a large number of previously seen objects. We have algorithms which deal only with recognizing shapes and algorithms which are suited for recognizing objects in cluttered scenes using shape, color and texture. This dissertation provides a unified framework which can be applied not only to recognize simple shapes such as silhouettes but also recognize real objects in cluttered environments with occlusion. The framework presented in this dissertation uses an accumulative approach reminiscent of the well known Generalized Hough Transform introduced by Ballard to recognize general shapes. Accumulator-based methods are highly parallel and use simple arithmetic. Noise and isotropic distortions tend to average out. The algorithm is invariant to translation, rotation, scale (zoom) and robust to illumination changes, background clutter, occlusion as well as view point changes. This is demonstrated using a wide range of data sets and experiments where it is shown to significantly outperform the current state-of-the-art. The novel contributions of this dissertation are as follows: 1. A unified matching algorithm which matches different object models by accumulating features in a higher dimensional space. 2. A general and unified object representation (or model) built using robust and invariant features, extracted based on the nature of the object.
Design and Development of a Cross Platform Interactive Image Processing System
(2005-11-28) Krish, Karthik; Dr. Wesley E. Snyder, Committee Chair; Dr. Joel Trussell, Committee Member; Dr. Edward Grant, Committee Member
The objective of this thesis is to design and develop a cross-platform software system which will enable the user to perform image processing/analysis interactively. The software system will have the ability to render images of any data-type and visualize them in many different ways. The software will also integrate many commonly used image processing and analysis algorithms, that can be run on the images. The cross-platform nature of the tool will help in making sure that a uniform interface is presented to the user irrespective of the underlying architecture. The system is designed to be modular which makes it suitable for future expansion.
Reconstruction of Lambertian Surfaces from Photometric Stereo
(2005-06-13) Sethuram, Amrutha Shree; Dr. Edward Grant, Committee Co-Chair; Dr. David Lalush, Committee Co-Chair; Dr. Wesley E. Snyder, Committee Chair
The objective of this thesis is to implement and compare two algorithms to reconstruct the shape of an object from photometric stereo. Photometric stereo is a practical technique for determining an object's shape and surface reflectance properties at a distance. The implementation proposes the use of three images of an object, recorded from the same viewpoint but with different illumination. The first algorithm employs the Fourier transform method to solve the minimization problem. The gradient data is obtained by incorporating photometric stereo method on image triplets. The Fourier transform of the unknown surface is then expressed as a function of the Fourier transform of the gradients. The relative depth values are then obtained by applying an inverse Fourier transform of the function. The second algorithm is based on iterative reconstruction which minimizes the cost function by gradient descent and annealing. Both these algorithms are implemented to reconstruct both real and synthetic surfaces and the results are compared. It is also shown that better reconstruction results are obtained by adopting the second algorithm in the presence of discontinuities in the image. Noise sensitivity of the frequency-domain method is also evaluated. An experimental setup to obtain real world images is also presented.