Optimization of the Jakim halal logo placement on qr code using enhanced genetic algorithm

• Main Author: Rohani, Fuaad
• Format: Thesis
• Language: English
• Published: 2018
• Subjects: QR codes - Case studies; Food - Religious aspects - Islam
• Online Access: http://psasir.upm.edu.my/id/eprint/76064/1/FK%202018%20147%20IR.pdf; http://psasir.upm.edu.my/id/eprint/76064/
• Institution: Universiti Putra Malaysia

A logo placed on Quick Response (QR) code provides additional aesthetic value and visual information to the user. Visual quality of the embedded logo is an important criterion besides success in QR code readability. Unfortunately, most work placed a logo on QR code with low embedded data capacity. Therefore, Jabatan Kemajuan Islam Malaysia (JAKIM) halal premises information is used to represent high data capacity. The thesis proposed a method to place a logo on top of the QR code, embedded with high data capacity, using Genetic Algorithm (GA) search technique to find appropriate size and location so the QR code can be decoded by various QR code decoders. Pad codewords modification technique is used to minimize the error introduced by the logo while three QR code decoders are used to achieve high probability of decode feasibility. A fitness function has been formulated to determine the appropriate size and location of the logo. A total of 3949 samples retrieved from JAKIM public access database was segregated into six groups based on the amount of information that could be embedded in the QR code, wherein 10% of the items from each group were used as samples for data generation. From the experiments, logo size about 69 to 76 pixels which covered about 5.39% to 7.16% of the QR code area can be used for all items in the respective group without decoding failure compared to 2% currently used in JAKIM Halal tag. The module pixel error in the QR code is found to be less than 4.25%. The logo placement system had successfully avoided error correction modules and control patterns, simultaneously, the placement location of the logo is maintained within the QR code area. The proposed work has been compared to GA-based technique, module modification technique and Simulated Annealing technique related to QR code readability, embedded logo visual quality and high embedded data capacity. As a conclusion, the system can successfully find a set of logo size and location on a QR code embedded with high data capacity that does not affect the QR code readability and can be decoded by various decoders with 100% accuracy.