Doctoral Dissertations
Date of Award
8-1995
Degree Type
Dissertation
Degree Name
Doctor of Philosophy
Major
Electrical Engineering
Major Professor
Mohan M. Trivedi
Committee Members
Mongi Abidi, Donald Bouldin, William Hamel, Marshall Pace
Abstract
Depth information is useful in many applications. The objective of this disser-tation is to develop a computational approach for extracting a three-dimensional structure of controllable resolution, depth of field, and accuracy at real-time speeds. The method combines the ability of stereo processing to acquire precise depth mea surements and the efficiency of a gradient based technique. Without any apriori information of the locations of the points in the scene, the correspondence problem in stereo processing is computationally expensive. In our approach, we develop a spatial and temporal gradient (STG) analysis, which has been shown to provide depth with high efficiency but limited accuracy, to guide the matching process of stereo. The STG approach utilizes the spatial and the temporal gradients of the streams of images acquired using an actively controlled camera. Depending on the require-ments of a particular task, appropriate parameters such as disparity value sought, interframe camera displacement, and number of frames in a stream, are chosen to control the resolution, depth of field, and accuracy. The camera motion used in the approach can be either lateral or axial. The acquisition and processing of the image stream are done in real-time on a pipeline architecture based processor. Extensive experiments are presented to demonstrate the accuracy, controllability of depth of field and resolution, and ability to perform successfully in a variety of scenes. The system operated with no latency between image acquisition and processing. The total acquisition and processing time in these experiments is in the range of 0.27 to 1.56 seconds. The depth results obtained using STG alone have an accuracy of 85 percent to 92 percent. The integration allows measurement of depth at a speed approximately ten-times higher than that of stereo processing alone. The extensive experiments on real-scenes have shown the ability to acquire depth with mean error of less than 3 percent.
Recommended Citation
Dalmia, Arun Kumar, "Depth measurements of selectable quality using image streams : computational framework for integrated spatio-temporal gradient analysis and binocular stereo approaches. " PhD diss., University of Tennessee, 1995.
https://trace.tennessee.edu/utk_graddiss/9967