UTILIZATION OF HYPERSPECTRAL IMAGING FOR CLASSIFYING POTATO TUBERS BASED ON ANTHOCYANIN CONTENT

Elmasry, G.; Radwan, S. M.; Wang, N.

doi:10.21608/jssae.2007.201259

UTILIZATION OF HYPERSPECTRAL IMAGING FOR CLASSIFYING POTATO TUBERS BASED ON ANTHOCYANIN CONTENT

Document Type : Original Article

Authors

¹ Agricultural Engineering Department, Faculty of Agriculture, Suez Canal University, Ismailia, Egypt.

² Department of Bioresource Engineering, McGill University, Macdonald Campus. 21,111 Lakeshore Road, Ste-Anne-de-Bellevue, Quebec. Canada. H9X 3V9.

10.21608/jssae.2007.201259

Abstract

Hyperspectral imaging technique in visible and near infrared (VIS/NIR) region with the range of 400 to 1000 nm was established for non-destructive assessment of anthocyanin concentration in three potato cultivars (Russet Burbank, Norland and Yokon Gold). Sixty tubers (20 tubers from each cultivar) were imaged with the hyperspectral imaging system and the anthocyanin concentrations were measured in the laboratory. Simple linear regression models were built between the anthocyanin concentration in the tested samples and their spectral responses at 550 nm. The prediction model performed well and predicted the anthocyanin content with standard error of calibration (SEC) equal to 7.22 in the calibration set (45 tubers) and correlation coefficient of 0.91. In the validation set (15 tubers), the model also achieved high success for predicting anthocyanin concentration with SEP of 11.95 and correlation coefficient of 0.91.

Keywords