Development of an AI-based Electronic nose and image processing techniques for detection of Botcha
No Thumbnail Available
Views
...
Downloads
...
Date
2023
Authors
King David O. Audencia and Clarizza Jane D. Pagani
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
The present investigation utilized artificial intelligence (AI)-based image processing and electronic nose sensing techniques to identify botcha meat. The findings of the investigation demonstrate the feasibility of detecting botcha through the utilization of two distinct techniques- the electronic nose technique and image processing technique .
The electronic nose technique relies on the specific presence of volatile compounds in botcha meat. Parameters such as vapor pressure, temperature, gas resistance and relative humidity were extracted and analyzed. With these parameters various combinations were established. Moreover, the combination of vapor pressure and temperature, vapor vapor pressure and gas gas resistance, and vapor pressure and relative humidity were observed to yield satisfactory results.
The image processing technique leverages the RGB color of meat. The mean of RGB color values, median of RGB color values and mode of RGB color values were extracted using image. The combinations of red mode value, blue mode value and green mode value was found to be especially significant.
In the present study, an AI-model was built using linear discriminant analysis and neutral network classification to accurately evaluate the detection of botcha. However, it has been observed that there is no statistically significant distinction between the performance of neutral network and linear discrimination analysis, no either of these analytical methods may be employed. There is no substantial distinction between electronic nose systems and methods of image analysis thus, e-nose and image processing technique can be utilized for the purpose of accurately detecting botcha meat.