Open AccessClassification of Milk Fish Quality using Fuzzy K-Nearest Neighbor Method Based on Form Descriptor and Co-Occurrence Matrix
Author(s) -
Deni Sutaji,
Rohman Dijaya
Publication year - 2019
Publication title -
journal of physics. conference series
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.21
H-Index - 85
eISSN - 1742-6596
pISSN - 1742-6588
DOI - 10.1088/1742-6596/1179/1/012021
Subject(s) - milkfish , quality (philosophy) , fishery , feature (linguistics) , fuzzy logic , fish products , computer science , fish <actinopterygii> , pattern recognition (psychology) , aquaculture , environmental science , business , artificial intelligence , biology , philosophy , linguistics , epistemology
Milkfish is a type of fish that lives in fresh water. Caught milkfish will be sorted at fish processing companies and will be a variety of milk, fish products in the form of cooked or in the form of fresh fish. A manual separation system of milkfish often experiences errors in the separation of fish species so that it takes a long time in the process of separating these types of fish, because it is very dependent on the skills of employees or employees. Errors in the separation of milkfish based on their size will cause a failure in the production section because it will affect the difference in the maturity level of a product, which will result in errors in the process of cooking the fish, which has an impact on the loss in the financial sector of a company. In order to maintain the quality of milkfish and speed up the sorting of milkfish, it is necessary to create a system that can help workers and employers in distinguishing the quality of milkfish based on their size. So the writer will create a system “Image Quality Classification of Milkfish Using Fuzzy K-Nearest Neighbor Method Based on Descriptors of Forms and Co-occurrence Testing was carried out using 45 images of milk fish. From the test results based on the shape feature and the co-assurance matrix produces 100%, which is in accordance with the form feature requirements and the co-insurance matrix as the reference data.