Tschannerl, Julius and Ren, Jinchang and Jack, Frances and Marshall, Stephen and Zhao, Huimin (2017) Employing NIR-SWIR hyperspectral imaging to predict the smokiness of Scotch whisky. In: Optical Characterization of Materials 2017, 2017-03-22 - 2017-03-23, Franhofer IOSB.
Text (Tschanner-etal-2017-Employing-NIT-SWIR-hyperspectral imaging-to-predict-the-smokiness)
Tschanner_etal_2017_Employing_NIT_SWIR_hyperspectral_imaging_to_predict_the_smokiness.pdf - Accepted Author Manuscript
Download (5MB) | Preview
Scotch Whisky makes a significant contribution to the UK's food and drinks export. The flavour of this high quality spirit is derived naturally from the whisky making process, with smoky aromas being a key character of certain Scotch whiskies. The level of smokiness is determined by the amount of phenolic compounds in the spirit. Phenols are introduced by exposing the barley malt to peat smoke during the kilning process. The current techniques to determine the levels of phenols, such as High Performance Liquid Chromatography (HPLC), are time consuming as they require distillation of the malt prior to analysis. To speed up this process and enable real-time detection before processing, the possibilities of Near-infrared to Short-wave-infrared (NIR-SWIR) Hyperspectral Imaging (HSI) to detect these phenols directly on malted barley are explored. It can be shown that via regression analysis, various levels of phenol concentration used as working levels for whisky production could be estimated to a satisfying degree. To further optimise industrial application, a hyperspectral band selection algorithm is applied that yields good results and reduces computational cost and may open possibilities to employ multispectral rather than hyperspectral cameras in future applications.
|Item type:||Conference or Workshop Item (Paper)|
|Keywords:||hyperspectral imaging , scotch whisky, near-infrared, band selection, Electrical engineering. Electronics Nuclear engineering, Food Science, Spectroscopy|
|Subjects:||Technology > Electrical engineering. Electronics Nuclear engineering|
|Department:||Faculty of Engineering > Electronic and Electrical Engineering
Technology and Innovation Centre > Sensors and Asset Management
|Depositing user:||Pure Administrator|
|Date Deposited:||28 Mar 2017 15:12|
|Last modified:||01 Apr 2017 02:36|