Accurate Real Time Detection For Halal Logo Based On Fourier Magnitude Method
The research works carried out is the analysis of object detection methods and development of a new method in recognizing the approved Halal logo by JAKIM and their implementation in Android device. Currently, many irresponsible entrepreneurs use imitation ‘Halal’ logo on their products. Consequent...
Saved in:
| Main Author: | |
|---|---|
| Format: | Thesis |
| Language: | English English |
| Published: |
2016
|
| Subjects: | |
| Online Access: | http://eprints.utem.edu.my/id/eprint/18179/1/Accurate%20Real%20Time%20Detection%20For%20Halal%20Logo%20Based%20On%20Fourier%20Magnitude%20Method%2024%20Pages.pdf http://eprints.utem.edu.my/id/eprint/18179/2/Accurate%20Real%20Time%20Detection%20For%20Halal%20Logo%20Based%20On%20Fourier%20Magnitude%20Method.pdf |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | The research works carried out is the analysis of object detection methods and development of a new method in recognizing the approved Halal logo by JAKIM and their
implementation in Android device. Currently, many irresponsible entrepreneurs use imitation ‘Halal’ logo on their products. Consequently, Muslim users find it hard to
determine the validity of ‘Halal’ logo used. This research aims to classify between JAKIM Halal logo and fake Halal logo. Beside that, the objective of this research is to develop an algorithm Fractionalized Principle Magnitude to recognize all 50 approved Halal logo. This research is divided into a three stages. In the first stage, the evaluations of the existing object detection methods in Android Smartphone is conducted. The evaluated object detection methods are Scale Invariant Feature Transform (SIFT), Speed up Robust Feature (SURF), Feature from Accelerate Segment Test (FAST), Good Feature to Track (GFTT), Maximally Stable Extremal Regions (MSER), Oriented FAST and Rotated BRIEF (ORB) and Center Surrounded Extrama (CenSurE). The characteristic of each object detection
method is studied and compared in order to identify the best object detection method that can be applied in the recognition of JAKIM Halal logo. The second stage is the recognition of JAKIM Halal logo using Android Smartphone. In this stage, object detection methods with good result from previous stage is evaluated and compared with a newly developed simple yet effective logo recognition method based template matching technique in recognizing JAKIM Halal logo from fake Halal logos on Android phones. The last stage is the final work to complete Malaysia Halal logo recognition system because JAKIM also approved other 50 Halal logo from around the world other than JAKIM Halal logo. So in the last stage, a novel logo recognition method based on Fourier magnitudes and k-nearest neighbor classifier is developed to recognize all 50 Halal logos that approved by JAKIM. This novel logo recognition method is called Fractionalized Principle Magnitude (FPM) and have been compared with other logo recognition method such as Histogram of Gradient (HOG), Hu Moment, Zernike Moment and Wavelet Co-occurrence Histogram (WCH). The comparison is carried out based on efficiency, consistency and accuracy
performances of each method. From the results, it shows that FPM obtains the highest average performance of 90.4% compared to those of 75.2% for HOG, 44.4% for Hu Moment, 64.4% for Zernike, and 47.2% for WCH. |
|---|