Doctor of Philosophy (PhD)
First Committee Member
Second Committee Member
Third Committee Member
Fourth Committee Member
Fifth Committee Member
Number of Pages
Most of robotic handling and assembly operations are based on sensors such as range and touch sensors. In certain circumstances, such as in the presence of ionizing radiation where most customary sensors will degrade over time due to radiation exposure, these sensors won't function properly. Utilizing two or more cameras (stereo vision) located outside the target zone and analyzing their images to identify location and dimensions of parts within the robot workspace is an alternative for using sensors. Object Recognition is affected by the light condition which oftentimes causes the gray-scale or red, green, and blue values to have a relatively small dynamic range. With this small dynamic range, edge detection algorithms fail to detect the proper edges and therefore cause improper image segmentation. To tackle this problem, a transformation on the (r,g,b) values of the pixels is introduced and applied prior to the edge detection and segmentation process. A stereoscopic computer vision system with multiple cameras is then used to compute the distance of the object from the origin of a global Euclidean coordinate system with high resolution. As an application of computer vision, a classifier for testing remote solar panels for cleanness condition, and performing cleaning when necessary, is introduced. A classification algorithm consisting of: the classification vector, the metric used, the training of the classifier, the testing of the classifier, and the classifier is put into play for everyday use. A smart cleaning robot is being designed based on this system to perform the cleaning autonomously when necessary. Another application of computer vision is inspecting the degree of air pollution. A real time classification algorithm that uses a quantization algorithm based on prior calibration is applied to evaluate the quality of air. The intelligent system, based on this algorithm, classifies the air using a numeric system from 1 to 10 which is then transformed to a qualitative scale.
Detectors; Pattern recognition systems; Radiation; Robotics; Robot vision; Robots
Mechanical Engineering | Robotics
University of Nevada, Las Vegas
Fayed, Ahmad A., "Computer-Based Stereoscopic Parts Recognition for Robotic Applications" (2013). UNLV Theses, Dissertations, Professional Papers, and Capstones. 1985.
IN COPYRIGHT. For more information about this rights statement, please visit http://rightsstatements.org/vocab/InC/1.0/