Classification of pork and beef meat images using extraction of color and texture feature by Grey Level Co-Occurrence Matrix method

3rd Annual Applied Science and Engineering Conference

工业技术;自然科学

Asmara, R.A.^1 ; Romario, R.^1 ; Batubulan, K.S.^1 ; Rohadi, E.^1 ; Siradjuddin, I.^2 ; Ronilaya, F.^2 ; Ariyanto, R.^1 ; Rahmad, C.^1 ; Rahutomo, F.^1

Informatics Engineering Department, State Polytechnic of Malang, Jl Soekarno Hatta 09, Malang, Indonesia^1

Electrical Engineering Department, State Polytechnic of Malang, Jl Soekarno Hatta 09, Malang, Indonesia^2

关键词: Cattle ranchers;    Color and texture features;    Financial loss;    First-order statistics;    Grey level co-occurrence matrixes;    Science and Technology;    Texture analysis;    Visual identification;   

Others  :  https://iopscience.iop.org/article/10.1088/1757-899X/434/1/012072/pdf
DOI  :  10.1088/1757-899X/434/1/012072

来源: IOP

【 摘 要 】

In Indonesia, beef meat usually sells with expensive price. It because of long distribution from the cattle rancher to the end customer. This causing a high-cost on the production and distribution. This condition made some people trying to be tricky by mixing beef meat with another meat, such as pork. It occurs the financial loss for customers. Currently, the meat identification done manually using visual identification of human vision. This method has a lot of weaknesses to differ beef and pork meat due because it need expert to identify the difference. The improvement of science and technology in computer vision helps the customer to identify beef and pork meat automatically using its images. This research classification of beef meat and pork using back-propagation method for its classifier. The purpose of this study is identifying differences in beef and pork based on digital image. Color feature using first order statistic mean from its RGB color value and using Grey Level Co-Occurrence Matrix for texture analysis. Experiment result of classification of beef and pork meat using Back-propagation, RGB color histogram value, and Grey Level Co-Occurrence Matrix achieving 89,57% of accuracy result.