Typescript (photocopy). / Digitized by Kansas Correctional Industries

Thesis (Ph. D.)--Industrial and Systems Engineering, Georgia Institute of Technology, 2007. / Yorai Wardi, Committee Member ; Gunter Sharp, Committee Member ; Spiridon Reveliotis, Committee Member ; Leon F. McGinnis, Committee Member ; Chen Zhou, Committee Chair.

Thesis submitted in compliance with the requirements for the Master's Degree in Technology: Industrial Engineering, Department of Industrial Engineering, Durban University of Technology, 2009. / This research focuses on the design and implementation of an intelligent machine vision and sorting system that can be used to sort objects in an industrial environment. Machine vision systems used for sorting are either geometry driven or are based on the textural components of an object’s image. The vision system proposed in this research is based on the textural analysis of pixel content and uses an artificial neural network to perform the recognition task. The neural network has been chosen over other methods such as fuzzy logic and support vector machines because of its relative simplicity. A Bluetooth communication link facilitates the communication between the main computer housing the intelligent recognition system and the remote robot control computer located in a plant environment. Digital images of the workpiece are first compressed before the feature vectors are extracted using principal component analysis. The compressed data containing the feature vectors is transmitted via the Bluetooth channel to the remote control computer for recognition by the neural network. The network performs the recognition function and transmits a control signal to the robot control computer which guides the robot arm to place the object in an allocated position. The performance of the proposed intelligent vision and sorting system is tested under different conditions and the most attractive aspect of the design is its simplicity. The ability of the system to remain relatively immune to noise, its capacity to generalize and its fault tolerance when faced with missing data made the neural network an attractive option over fuzzy logic and support vector machines.

Thesis submitted in compliance with the requirements for the Master's Degree in Technology: Industrial Engineering, Department of Industrial Engineering, Durban University of Technology, 2009. / This research focuses on the design and implementation of an intelligent machine vision and sorting system that can be used to sort objects in an industrial environment. Machine vision systems used for sorting are either geometry driven or are based on the textural components of an object’s image. The vision system proposed in this research is based on the textural analysis of pixel content and uses an artificial neural network to perform the recognition task. The neural network has been chosen over other methods such as fuzzy logic and support vector machines because of its relative simplicity. A Bluetooth communication link facilitates the communication between the main computer housing the intelligent recognition system and the remote robot control computer located in a plant environment. Digital images of the workpiece are first compressed before the feature vectors are extracted using principal component analysis. The compressed data containing the feature vectors is transmitted via the Bluetooth channel to the remote control computer for recognition by the neural network. The network performs the recognition function and transmits a control signal to the robot control computer which guides the robot arm to place the object in an allocated position. The performance of the proposed intelligent vision and sorting system is tested under different conditions and the most attractive aspect of the design is its simplicity. The ability of the system to remain relatively immune to noise, its capacity to generalize and its fault tolerance when faced with missing data made the neural network an attractive option over fuzzy logic and support vector machines.

Oyster meats are currently sorted by hand using volume as the sorting parameter. Hand grading is inaccurate, time consuming and costly. Previous research on physical properties of oyster meats showed a high correlation between projected area of oyster meats and their volume thus allowing the use of projected area measurements as a sorting criterion. A machine vision based oyster meat sorting machine was developed to mechanize the sorting process. The machine consists of a dark conveyor belt transporting singulated oysters through a grading station and then along a row of fast acting water jet valves which separates the stream of oysters into 3 classes. The vision system consists of a monochrome television camera, flash light illumination to "freeze" the images, a digitizer/transmitter and a Personal Computer as an image processing unit. Software synchronizes the flash light and digitization of images and calculates projected area of each meat using the planimeter method. The grading results are sent to a valve control board which actuates the spray valves. The sorting rate is 37 oyster meats/min with a sorting accuracy of 87.5%. A description of the design work, adjustment and l calibration procedures and a final sorting test is included. / Master of Science

The design and operation of a high volume conveyor sortation system are important due to its high cost, large footprint and critical role in the system. In this thesis, we study the characteristics of the conveyor sortation system from performance evaluation and design perspectives employing continuous modeling approaches. We present two continuous conveyor models (Delay and Stock Model and Batch on Conveyor Model) with different representation accuracy in a unified mathematical framework. Based on the Batch on Conveyor Model, we develop a fast fluid simulation methodology. We address the feasibility of implementing fluid simulation from modeling capabilities, algorithm design and simulation performance in terms of accuracy and simulation time. From a design perspective, we focus on rates determination and accumulation design in the accumulation and merge subsystem. The optimization problem is to find a minimum cost design that satisfies some predefined performance requirements under stochastic conditions. We first transform this stochastic programming problem into a deterministic nonlinear programming problem through sample path based optimization method. A gradient based method is adopted to solve the deterministic problem. Since there is no closed form for performance metric even for a deterministic input stream, we adopt continuous modeling to develop deterministic performance evaluation models and conduct sensitivity analysis on these models. We explore the prospects of using the two continuous conveyor models we presented.

This report discusses in great detail the various research, design, and development stages of the Produce Sorting Microcomputer developed for HAGAN ENGINEERING Inc. The two-semester Clinical Research project has been approved by the graduate committee at the School of Engineering at the University of the Pacific and fulfills the requirements towards a Master Degree in Electrical Engineering. The project was selected based on its complexity, feasibility, the time span it required to complete, and its relevance to the area of real time microcomputer design. In addition, the design constraints and specifications were to be dictated solely by HAGAN ENGINEERING Inc. and all further modifications were to be discussed and approved by HAGAN. These limitations created a professional industry-like atmosphere, which is one of the goals of the Clinical Research Program. A brief User's Manual will accompany the MC68000 board; it will contain all the vital information about the system that a programmer or a technician might need to understand the system. The manual wall contain the complete circuit schematic, a parts list, general design features, and all the software properties of the system (memory map, interrupt tables register map).