Engram, J. Todd (2007) Innovative approach to control of an emulsion loader. Masters thesis, Memorial University of Newfoundland.
- Accepted Version
Available under License - The author retains copyright ownership and moral rights in this thesis. Neither the thesis nor substantial extracts from it may be printed or otherwise reproduced without the author's permission.
In this thesis, the work to develop the vision system for the Emulsion Loading Automation Project (ELAP) is presented. The ELAP project was a collaborative effort among various team members to develop an intelligent system to automate the complex task of an emulsion loader used in underground mining operations. The vision system was tasked with the challenging imaging requirements of locating and identifying drill-holes in the rock face of an underground mine, and to provide visual guidance for the hose guide to load the drill-holes with emulsion. The first part of this thesis outlines the development of image processing algorithms to segment and identify drill-holes present in an image acquired from a video stream. This involved applying thresholding and morphological techniques to preprocess the image to improve contrast, separate touching objects, and fill any holes. A pattern recognition model was then developed using drill-hole feature data to classify segmented objects as either drill-holes or background artifacts. The second part of this thesis presents the work performed for visual guidance to position the robotic boom in front of a drill-hole for loading. Using a camera mounted to the end of a robotic boom, camera and hand-eye calibration routines were developed to transform the drill-hole image object to both the camera and end-effector reference frames. A visual guidance algorithm was then developed using the calibration parameters to visual servo the robotic boom to a drill-hole for loading. Finally, testing performed in an underground mine after the critical subsystems were integrated and operational, verified the vision system operation as designed.
|Item Type:||Thesis (Masters)|
|Additional Information:||Includes bibliographical references (leaves 115-120).|
|Department(s):||Engineering and Applied Science, Faculty of|
|Library of Congress Subject Heading:||Computer vision; Image processing; Mineral industries--Technological innovations; Mining engineering--Automation.|
Actions (login required)