DIFERENSIASI KUAH TERKONTAMINASI MINYAK BABI DAN MINYAK SAPI BERBASIS NILAI RGB MENGGUNAKAN HIGH POWER UV-LED FLUORESCANCE IMAGING SYSTEM TERKOMBINASI MACHINE LEARNING BERALGORITMA K-NEAREST NEIGHBOR
| Main Author: | Khoirini Mawaddah, NIM.: 18106020026 |
|---|---|
| Format: | Thesis NonPeerReviewed Book |
| Bahasa: | ind |
| Terbitan: |
, 2022
|
| Subjects: | |
| Online Access: |
https://digilib.uin-suka.ac.id/id/eprint/56139/1/18106020026_BAB-I_IV-atau-V_DAFTAR-PUSTAKA.pdf https://digilib.uin-suka.ac.id/id/eprint/56139/2/18106020026_BAB-II_sampai_SEBELUM-BAB-TERAKHIR.pdf https://digilib.uin-suka.ac.id/id/eprint/56139/ |
Daftar Isi:
- The background of this research was fraudulent mixing of lard in the broth, while the current test method (RT-PCR) required professional staff to operate and the testing costs are expensive. This study aimed to detect and differentiate RGB values in gravy images contaminated with pork oil and cow oil using a high power UV-LED fluorescence imaging system combined with machine learning with the K-Nearest Neighbor algorithm. This research was conducted in three stages, namely samples making, data collection, and data processing. The samples in this study were 10 cups of broth contaminated with pork oil and cow oil. Data collection was carried out by detecting samples of broth contaminated with lard and cow oil using a high power UV-LED fluorescence imaging system so that 100 RGB values of both were obtained. Data processing was carried out using Machine Learning with the K-NN algorithm created using RapidMiner software. The results showed that broths contaminated with pork oil and cow oil were successfully detected using a high power UV-LED fluorescence imaging system and differentiated using a machine learning K-NN algorithm with very good quality with 100% accuracy, 100% precision and recall, and AUC value of 1.0.