Tan , Jin Siang
(2017)
Detection Of Misplaced And Missing Regions In Image Using Neural Network.
Masters thesis, Universiti Sains Malaysia.
Abstract
Jigsaw puzzle is a printed picture that is cut into various pieces of different shapes. The game requires assembly of many oddly shaped pieces into producing a complete picture. However, misplaced or missing jigsaw puzzle pieces are difficult to be detected by human eyes. This scenario can be bridged to circuit on the breadboard, which also has similar condition like having several components on the board. Based on research, most of the algorithms are not intelligent enough and only able to detect the missing component. Therefore, it is necessary to develop an algorithm that is able to detect both misplaced and missing jigsaw puzzles. The main objective of this project is to develop an intelligent system to solve the jigsaw puzzle using Matlab software. The developed system consists of the image processing and the neural network phases. In image processing phase, the captured image is split into regions and the RGB (Red Green Blue) value of the regions is obtained. The neural network used in this research is a back-propagation neural network and it is trained by using Scaled Conjugate Gradient training algorithm. The neural network uses the RGB value from the image processing phase and analyzes the regions to check whether there is misplaced or missing jigsaw puzzle. Two experiments have been conducted, which are time performance in order for the system to analyze the regions and the ability of the system in detecting the misplaced and missing jigsaw puzzle. From the result, it is found that the time needed for the system to analyze 20 pieces of the image is around 89 seconds. The system also gives almost 100% of accuracy in detecting the missing or misplaced regions of the jigsaw puzzles image.
Actions (login required)
 |
View Item |
![[img]](http://eprints.usm.my/style/images/fileicons/application_pdf.png)
