Salt Content Prediction System of Dried Sea Cucumber (Beche-de-mer) Based on Visual Near-Infrared Imaging

Abstract

Dried sea cucumber (Beche-de-mer) is a culinary food that is considered luxurious and delicious, especially in China, Korea, and Japan, so the price is quite high. Dried sea cucumber (Beche-de-mer) also has high commercial value and high nutritional value. Their quality determines dried sea cucumber (Beche-de-mer) prices on international markets. One of the parameters that determine its quality is salt content. The excessive salt content in Dried sea cucumber (Beche-de-mer) can cause health problems such as hypertension, stroke, digestive system disorders, etc. Therefore, this paper will discuss a prediction system for measuring salt content in Dried sea cucumber (Beche-de-mer) using hyperspectral imaging technique. This system uses reflectance mode with a wavelength from 400 to1000 nm. The hardware from the prediction system for measuring salt content is motors to generate, hyperspectral camera system, two 150 W halogen lamps, Teflon tables, and personal computer link. Then, the PLSR algorithm is applied to the prediction system model at full wavelength. The prediction model is used to obtain the predicted value of salt content. Then the results of the prediction model are compared with the data references obtained by the mercury nitrate method. The root means square errors and correlation coefficient are used to evaluate the prediction system performance of salt content. The best result of the prediction system in this work is to have a correlation coefficient of 0.99 and root mean square errors of 0.27, respectively, with the number of PLS component is 25. Based on the results of this work, the proposed system can be used as an alternative method of measuring the salt content in dried sea cucumber (Beche-de-mer) with excellent accuracy and high reliability.