DefinePK hosts the largest index of Pakistani journals, research articles, news headlines, and videos. It also offers chapter-level book search.
Title: Fish quality study using Odor-Profile Case-Based Reasoning (CBR) classification technique
Authors: Muhammad Sharfi Najib, Nurul Hafizah Zamberan, Nurdiyana Zahed, Fathimah Abdul Halim, Muhammad Faruqi Zahari, Wan Muhamad Azmi Mamat, Hadi Manap
Journal: ARPN Journal of Engineering and Applied Sciences
Publisher: Khyber Medical College, Peshawar
Country: Pakistan
Year: 2016
Volume: 11
Issue: 10
Language: English
Fish has high quality protein and other essential nutrients and are an important part of a healthful diet. It is important to make sure the quality of fish to avoid food poisoning. There are three methods to know the quality of fish which is sensory, microbiology and chemical methods. Nowadays, some fish monger use formalin to make fish looks fresh and good. Formalin is a colorless strong-smelling chemical substance usually used in industry of textiles, plastics, papers, paint, construction, and well known to preserve human corpse. It is derived from formaldehyde gas dissolved in water. Exposure to the gas and vapour can make irritation to the eyes, nose and respiratory tract. There is quite difficult for consumers to differentiate fresh fish without formalin and fish with formalin. This is because, they look fresh and good but the different is the odor which is the fresh fish still have a fishy smell while fish with formalin do not smelly. Therefore, we use electronic nose (E-nose) to know fresh fish and formalin-based preserved fish. E-nose consists of an array of conduct metric chemical sensors which change resistance when exposed to vapour. The odor-profile of the fish samples were collected based on designated experimental procedure. The measured raw data was then stored in Microsoft Excel data and converted into MATLAB. Then they were normalized and their unique features were extracted using statistical tools. The input features were than inserted into Case-based Reasoning (CBR) library and intelligently classified using CBR method and validated based specific performance measure. The results have shown that the CBR classified with 100.00% rate of accuracy.
Loading PDF...
Loading Statistics...