Wyniki wyszukiwania - BazTech

Ograniczanie wyników

Znaleziono wyników: 1

Liczba wyników na stronie

Wyniki wyszukiwania

Wyszukiwano:
w słowach kluczowych: grain infestation

Sortuj według:

Ogranicz wyniki do:

Development of a method for assessing the quality of barley for brewing using hyperspectral imaging

Zapotoczny Piotr

Technical Sciences / University of Warmia and Mazury in Olsztyn

2024

Vol. 27

357--375

The paper discusses the utilization of hyperspectral imaging in the process of assessing the quality of barley grain intended for brewing purposes. A specialized research setup comprising a spectrophotometer coupled with a CCD camera was employed. During measurements, the spectral distribution of each pixel in the image was recorded within the range of 400 to 1000 nm, enabling the extraction of homogeneous areas on the grain surfaces. Subsequently, surface texture parameters were computed on the designated areas. Prior to engaging in classification analyses, variable reduction was performed utilizing: (a) Fisher's coefficient, (b) classification error coefficient along with the averaged correlation coefficient POE+ACC, and (c) mutual information coefficient MI. The research material consisted of grain categorized into rain-soaked (B), mold -infested (M), and healthy (H). The best classification results were obtained for a wavelength of 800 nm from the extracted homogeneous areas. The classification accuracy reached 100% across all quality groups.