A genetic algorithm to obtain optimum parameters for a halcon vision system
- Authors: Fulton, Dale Meares
- Date: 2017
- Subjects: Genetic algorithms , Artificial intelligence , Automation , User interfaces (Computer systems)
- Language: English
- Type: Thesis , Masters , MEng
- Identifier: http://hdl.handle.net/10948/29751 , vital:30774
- Description: This report discusses the optimisation of a HALCON vision system using artificial intelligence, specifically a genetic algorithm. Within industrial applications, vision systems are often used for automated part inspection and quality control. A number of vision system parameters are to be selected when setting up a vision system. Since each vision system application differs, there is no specific set of optimal parameters. Parameters are selected during installation using a trial and error method. As a result, there is a need for an automated process for obtaining suitable vision system parameters. Within this report, research was conducted on both vision systems, genetic algorithms and integration of the two. A physical vision system was designed and developed utilising HALCON vision software. A genetic algorithm was then developed and integrated with the vision system. After integration, experimental testing was performed on the genetic algorithm in order to determine the ideal genetic algorithm control parameters which yield ideal genetic algorithm performance. Once the ideal genetic algorithm was obtained, the genetic algorithm was applied to the vision system in order to obtain optimal vision system parameters. Results showed that applying the genetic algorithm to the vision system optimised the vision system performance well.
- Full Text:
- Date Issued: 2017
- Authors: Fulton, Dale Meares
- Date: 2017
- Subjects: Genetic algorithms , Artificial intelligence , Automation , User interfaces (Computer systems)
- Language: English
- Type: Thesis , Masters , MEng
- Identifier: http://hdl.handle.net/10948/29751 , vital:30774
- Description: This report discusses the optimisation of a HALCON vision system using artificial intelligence, specifically a genetic algorithm. Within industrial applications, vision systems are often used for automated part inspection and quality control. A number of vision system parameters are to be selected when setting up a vision system. Since each vision system application differs, there is no specific set of optimal parameters. Parameters are selected during installation using a trial and error method. As a result, there is a need for an automated process for obtaining suitable vision system parameters. Within this report, research was conducted on both vision systems, genetic algorithms and integration of the two. A physical vision system was designed and developed utilising HALCON vision software. A genetic algorithm was then developed and integrated with the vision system. After integration, experimental testing was performed on the genetic algorithm in order to determine the ideal genetic algorithm control parameters which yield ideal genetic algorithm performance. Once the ideal genetic algorithm was obtained, the genetic algorithm was applied to the vision system in order to obtain optimal vision system parameters. Results showed that applying the genetic algorithm to the vision system optimised the vision system performance well.
- Full Text:
- Date Issued: 2017
Genetic algorithm for Artificial Neural Network training for the purpose of Automated Part Recognition
- Authors: Buys, Stefan
- Date: 2012
- Subjects: Genetic algorithms , Software architecture
- Language: English
- Type: Thesis , Masters , MSc
- Identifier: vital:9648 , http://hdl.handle.net/10948/d1008356 , Genetic algorithms , Software architecture
- Description: Object or part recognition is of major interest in industrial environments. Current methods implement expensive camera based solutions. There is a need for a cost effective alternative to be developed. One of the proposed methods is to overcome the hardware, camera, problem by implementing a software solution. Artificial Neural Networks (ANN) are to be used as the underlying intelligent software as they have high tolerance for noise and have the ability to generalize. A colleague has implemented a basic ANN based system comprising of an ANN and three cost effective laser distance sensors. However, the system is only able to identify 3 different parts and needed hard coding changes made by trial and error. This is not practical for industrial use in a production environment where there are a large quantity of different parts to be identified that change relatively regularly. The ability to easily train more parts is required. Difficulties associated with traditional mathematically guided training methods are discussed, which leads to the development of a Genetic Algorithm (GA) based evolutionary training method that overcomes these difficulties and makes accurate part recognition possible. An ANN hybridised with GA training is introduced and a general solution encoding scheme which is used to encode the required ANN connection weights. Experimental tests were performed in order to determine the ideal GA performance and control parameters as studies have indicated that different GA control parameters can lead to large differences in training accuracy. After performing these tests, the training accuracy was analyzed by investigation into GA performance as well as hardware based part recognition performance. This analysis identified the ideal GA control parameters when training an ANN for the purpose of part recognition and showed that the ANN generally trained well and could generalize well on data not presented to it during training.
- Full Text:
- Date Issued: 2012
- Authors: Buys, Stefan
- Date: 2012
- Subjects: Genetic algorithms , Software architecture
- Language: English
- Type: Thesis , Masters , MSc
- Identifier: vital:9648 , http://hdl.handle.net/10948/d1008356 , Genetic algorithms , Software architecture
- Description: Object or part recognition is of major interest in industrial environments. Current methods implement expensive camera based solutions. There is a need for a cost effective alternative to be developed. One of the proposed methods is to overcome the hardware, camera, problem by implementing a software solution. Artificial Neural Networks (ANN) are to be used as the underlying intelligent software as they have high tolerance for noise and have the ability to generalize. A colleague has implemented a basic ANN based system comprising of an ANN and three cost effective laser distance sensors. However, the system is only able to identify 3 different parts and needed hard coding changes made by trial and error. This is not practical for industrial use in a production environment where there are a large quantity of different parts to be identified that change relatively regularly. The ability to easily train more parts is required. Difficulties associated with traditional mathematically guided training methods are discussed, which leads to the development of a Genetic Algorithm (GA) based evolutionary training method that overcomes these difficulties and makes accurate part recognition possible. An ANN hybridised with GA training is introduced and a general solution encoding scheme which is used to encode the required ANN connection weights. Experimental tests were performed in order to determine the ideal GA performance and control parameters as studies have indicated that different GA control parameters can lead to large differences in training accuracy. After performing these tests, the training accuracy was analyzed by investigation into GA performance as well as hardware based part recognition performance. This analysis identified the ideal GA control parameters when training an ANN for the purpose of part recognition and showed that the ANN generally trained well and could generalize well on data not presented to it during training.
- Full Text:
- Date Issued: 2012
A general genetic algorithm for one and two dimensional cutting and packing problems
- Authors: Mancapa, Vusisizwe
- Date: 2007
- Subjects: Packaging -- Data processing , Genetic algorithms , Cutting stock problem , Packing for shipment , Manufacturing processes -- Planning
- Language: English
- Type: Thesis , Masters , MTech
- Identifier: vital:9602 , http://hdl.handle.net/10948/555 , http://hdl.handle.net/10948/d1011727 , Packaging -- Data processing , Genetic algorithms , Cutting stock problem , Packing for shipment , Manufacturing processes -- Planning
- Description: Cutting and packing problems are combinatorial optimisation problems. The major interest in these problems is their practical significance, in manufacturing and other business sectors. In most manufacturing situations a raw material usually in some standard size has to be divided or be cut into smaller items to complete the production of some product. Since the cost of this raw material usually forms a significant portion of the input costs, it is therefore desirable that this resource be used efficiently. A hybrid general genetic algorithm is presented in this work to solve one and two dimensional problems of this nature. The novelties with this algorithm are: A novel placement heuristic hybridised with a Genetic Algorithm is introduced and a general solution encoding scheme which is used to encode one dimensional and two dimensional problems is also introduced.
- Full Text:
- Date Issued: 2007
- Authors: Mancapa, Vusisizwe
- Date: 2007
- Subjects: Packaging -- Data processing , Genetic algorithms , Cutting stock problem , Packing for shipment , Manufacturing processes -- Planning
- Language: English
- Type: Thesis , Masters , MTech
- Identifier: vital:9602 , http://hdl.handle.net/10948/555 , http://hdl.handle.net/10948/d1011727 , Packaging -- Data processing , Genetic algorithms , Cutting stock problem , Packing for shipment , Manufacturing processes -- Planning
- Description: Cutting and packing problems are combinatorial optimisation problems. The major interest in these problems is their practical significance, in manufacturing and other business sectors. In most manufacturing situations a raw material usually in some standard size has to be divided or be cut into smaller items to complete the production of some product. Since the cost of this raw material usually forms a significant portion of the input costs, it is therefore desirable that this resource be used efficiently. A hybrid general genetic algorithm is presented in this work to solve one and two dimensional problems of this nature. The novelties with this algorithm are: A novel placement heuristic hybridised with a Genetic Algorithm is introduced and a general solution encoding scheme which is used to encode one dimensional and two dimensional problems is also introduced.
- Full Text:
- Date Issued: 2007
- «
- ‹
- 1
- ›
- »