PENGENALAN KARAKTER AKSARA JAWA MENGGUNAKAN KOMPUTASI PARAREL PADA SEGMENTASI CITRA

Nugroho, Oskar Ika Adi (2015) PENGENALAN KARAKTER AKSARA JAWA MENGGUNAKAN KOMPUTASI PARAREL PADA SEGMENTASI CITRA. S2 thesis, UAJY.

[img] Text (Halaman Judul)
MTF0010611.pdf

Download (208kB)
[img] Text (Bab I)
MTF1010611.pdf

Download (60kB)
[img] Text (Bab II)
MTF2010611.pdf

Download (300kB)
[img] Text (Bab III)
MTF3010611.pdf
Restricted to Registered users only

Download (57kB)
[img] Text (Bab IV)
MTF4010611.pdf
Restricted to Registered users only

Download (295kB)
[img] Text (Bab V)
MTF5010611.pdf
Restricted to Registered users only

Download (517kB)
[img] Text (Bab VI)
MTF6010611.pdf

Download (129kB)

Abstract

Aksara Jawa (java script) is a characteristic and heritage of the Javanese who need to be protected and preserved its existence. The author develops software digital image character recognition of Aksara Jawa (java script). In the process of character recognition, image segmentation process is required. The purpose of image segmentation is done to change the representation of an image into something more meaningful and easier to analyze. Image segmentation using clustering methods can be used with a variety of methods, one of which is the Particle Swarm Optimization (PSO). In the process used method of Particle Swarm Optimization (PSO) for the segmentation process in Aksara Jawa (java script) image, and using back propagation neural network for character recognition process of Aksara Jawa (java script). To overcome the computational burden is quite heavy in the process of image segmentation, then diterapakanlah method of parallel programming, and computer architectures that support CUDA GPU from NVIDIA. The end result of this research is to implement a digital image segmentation method of Particle Swarm Optimization in Aksara Jawa (java script) and implement the backpropagation method of training a neural network to perform Aksara Jawa (java script) character recognition on a digital image of Aksara Jawa (java script). CUDA GPU accelerating the process of image segmentation Aksara Jawa (java script). The average percentage of each character recognition accuracy of the results for the same font with training reached 75%.

Item Type: Thesis (S2)
Uncontrolled Keywords: aksara jawa, segmentation, character recognition, particle swarm optimization, backpropagation neural network, parallel programing, GPU, CUDA.
Subjects: Magister Teknik Informatika > Soft Computing
Divisions: Pasca Sarjana > Magister Teknik Informatika
Depositing User: Editor UAJY
Date Deposited: 26 Jun 2015 09:18
Last Modified: 26 Jun 2015 09:18
URI: http://e-journal.uajy.ac.id/id/eprint/7523

Actions (login required)

View Item View Item