Click on “Download PDF” for the PDF version or on the title for the HTML version.

If you are not an ASABE member or if your employer has not arranged for access to the full-text, Click here for options.

Moisture content measurement in dried apple produce through visible wavelength hyperspectral imaging

Published by the American Society of Agricultural and Biological Engineers, St. Joseph, Michigan

Citation:  2015 ASABE Annual International Meeting  152186400.(doi:10.13031/aim.20152186400)
Authors:   Stuart Crichton, Barbara Sturm, Anya Hurlbert
Keywords:   hyperspectral imaging, apples, moisture content, drying, image processing, PLSR.

Abstract. Traditional fruit drying within industry is undertaken using tried, tested and trusted drying schemes for different food produce. This maximises yield for a specific output food quality when input food quality is strictly controlled. However this strict control of input food quality leads to excess food wastage at the input stage. The use of set drying treatments can also result in higher energy usage during the drying process than may be necessary. In order to investigate methods to reduce both food wastage and energy usage we propose the tailoring of the drying process to the incoming food quality and condition.To achieve this we investigate the feasibility of online moisture content estimation of apple discs during the drying process through the use of hyperspectral imaging in the visible range (400-780nm). This involved the drying of apple samples, Braeburn variety, within a general purpose oven, Genlab J-SS-OH, at 70°C. Imaging was undertaken using a Specim V10E hyperspectral camera with a rotational scanning mirror attachment and Schneider Xenoplan 50mm f/2.8 lens under illumination provided by a Verivide D65 ‘Artificial Daylight’ lamp (CCT 6500K), in a viewing cabinet. 54 Braeburn apples, at the same ripeness stage, were sliced into 5mm thick discs of uniform size and inner and outer diameters. These were then dried within the oven at 70°C for a period of 6 hours. Each disc was individually hyperspectrally imaged at 1 hour intervals in conjunction with mass measurement of each apple disc. From regression analysis, a good level of moisture content prediction within this wavelength region was uncovered.

(Download PDF)    (Export to EndNotes)