SITEKIN: Jurnal Sains, Teknologi dan Industri
Vol 14, No 2 (2017): JUNI 2017

ANALISA METODE GABOR DAN PROPBABILISTIC NEURAL NETWORK UNTUK KLASIFIKASI CITRA (STUDI KASUS: CITRA DAGING SAPI DAN BABI)

Lestari Handayani (UIN Sultan Syarif Kasim Riau)



Article Info

Publish Date
11 Jun 2017

Abstract

Research for image classification can be utilized by Muslims to recognize the image of halal or haram meat. In this study, Gabor method analysis for image classification, especially distinguishing the image of beef and pork as an example of halal and unlawful meat. The image of the beef, the image of pork and the image of the beef and pork mixed. The image was obtained from several markets in Pekanbaru. The number of beef was 200 images, 100 images of pork and 50 images of both mixed. The system was built using PHP-based web programming language that can be accessed by the public. This meat image classification system produces an image output known as beef or pork. For the image of beef and pork mixed expected to be recognized as pork. But from the experiments that use 30% of the total image as the test data obtained the result that the accuracy of image classification reached 57.14%. And for image mixed obtain 46.67% recognized as pork. This is because the image of beef and pork is very similar when viewed from the texture only. This research can still be enhanced by adding other characters feature extraction methods.

Copyrights © 2017






Journal Info

Abbrev

sitekin

Publisher

Subject

Control & Systems Engineering Decision Sciences, Operations Research & Management Economics, Econometrics & Finance Industrial & Manufacturing Engineering Other

Description

Sesuai dengan standard ISO 45001 bahwa karyawan harus berpartisipasi dalam melakukan pencegahan kecelakaan. Untuk itu perusahaan telah menetapkan Program Hazob (Hazard Observation) untuk mengidentifikasi bahaya dan melakukan tindakan koreksinya. Penerapan Program Hazob masih dengan metode ...