Image processing on intel edison microcontroller for pork detection

• Main Author: Khdhair, Zaid Hadi
• Format: Thesis
• Language: English
• Published: 2016
• Online Access: http://psasir.upm.edu.my/id/eprint/70535/1/FK%202016%20104%20IR.pdf; http://psasir.upm.edu.my/id/eprint/70535/

Around 0.5% of females and 8% of males are suffer from color vision deficiency (CVD) problem. Moreover, CVD becomes more critical in case of professional laboratory operators, including halal laboratories in terms of the need of visual inspection. Halal and haram are the two terms used to indicate the allowed and not allowed food for Muslims respectively. Recently, the researchers have demonstrated that it is essential to develop inexpensive, and reliable methods to support halal food detection systems. Hence, detecting pork meat cheating in beef and chicken meatballs visually has been researched by using gold nanoparticles (GNPs) which changes the color of pork DNA samples into a clearly remarkable color. The result of GNPs research cannot be reliable due to CVD errors, thus, spectroscopy device has been used to handle this issue. However, using spectroscopy device to solve CVD has various disadvantages including high cost, and it should be used immediately after preparing the chemical solution in which it returns back to its original color after few minutes. In this research, implementing image processing techniques based on Intel Edison platform is used to handle CVD problem and decrease the detection cost. Moreover, image enhancement techniques have been used to label the color type of GNPs research. Color balance and brightness adjustments are applied for image enhancement aim, and the color type has been identified based on its intensity value. Python programming language and its Python Imaging Library (PIL) have been employed to implement the image processing tasks. On the other hand, Intel Edison and a traditional computer have implemented the Python algorithm, and the performance of each system has been recorded. The proposed work has been evaluated and compared to the performance of the traditional computer and to the result of absorption spectroscopy device. The result of this research has indicated that, image processing techniques, which are based on Intel Edison, are efficient alternative method to the absorption spectroscopy device, since both techniques have identified color types and solved CVD. Moreover, the proposed method is cost effective, and it can be used anytime since it is based on digital data. Thus, the proposed work can be used for future halal system designs, due to the tiny size (60×29×8 mm), light weight (32 g), low cost, and sufficient processing speed (500MHz Dual-core) of Intel Edison, and the promising results of image processing techniques.