Nondestructive detection of multiple dried squid qualities by hyperspectral imaging combined with 1D-KAN-CNN
Abstract: Given that dried squid is a highly regarded marine product in Oriental countries, the global food industry requires a swift and noninvasive quality assessment of this product. The current study therefore uses visiblenear-infrared (VIS-NIR) hyperspectral imaging and deep learning (DL) methodologies. We acquired and preprocessed VIS-NIR (4001000 nm) hyperspectral reflectance images of 93 dried squid samples. Important wavelengths were selected using competitive adaptive reweighted sampling, principal component analysis, and the successive projections algorithm. Based on a Kolmogorov-Arnold network (KAN), we introduce a one-dimensional, KAN convolutional neural network (1D-KAN-CNN) for nondestructive measurements of fat, protein, and total volatile basic nitrogen….
Tags: academic papers, squid