Deep and Machine Learning Using SEM, FTIR, and Texture Analysis to Detect Polysaccharide in Raspberry Powders

Abstract

In the paper, an attempt was made to use methods of artificial neural networks (ANN) and Fourier transform infrared spectroscopy (FTIR) to identify raspberry powders that are different from each other in terms of the amount and the type of polysaccharide. Spectra in the absorbance function (FTIR) were prepared as well as training sets, taking into account the structure of microparticles acquired from microscopic images with Scanning Electron Microscopy (SEM). In addition to the above, Multi-Layer Perceptron Networks (MLPNs) with a set of texture descriptors (machine learning) and Convolution Neural Network (CNN) with bitmap (deep learning) were devised, which is an innovative attitude to solving this issue. The aim of the paper was to create MLPN and CNN neural models, which are characterized by a high efficiency of classification. It translates into recognizing microparticles (obtaining their homogeneity) of raspberry powders on the basis of the texture of the image pixel.

Physicochemical and Morphological Study of the Saccharomyces cerevisiae Cell-Based Microcapsules with Novel Cold-Pressed Oil Blends

“…FT-IR spectra were obtained via a spectrophotometer of the Perkin Elmer company (Waltham, MA, USA), equipped with an ATR device with diamond as the internal reflection element [26]. Data were collected over a spectral range of 4000-400 cm −1 .…”

Guidelines for a Morphometric Analysis of Prokaryotic and Eukaryotic Cells by Scanning Electron Microscopy

“…The use of these descriptors is usually called “perceptual characterisation” and is useful for a coarse classification of textures [ 98 ]. A quantitative description of a texture requires the use of advanced computational methods, for instance a multiresolution decomposition using Gabor wavelets [ 98 ], artificial neural networks [ 99 ], as well as a textural analysis based on a gray-level co-occurrence matrix (GLCM) [ 100 ].…”

“…Since SEM enables to capture fine shapes of biological objects, this technique has been routinely used to collect images suitable for the analysis of textural descriptors [ 99 , 100 , 101 , 102 ]. Still, it should be borne in mind that the processing method could affect the texture of a sample.…”

“…To perform a quantitative analysis, however, a high number of SEM micrographs should be collected and processed point by point [ 99 ]. The most popular texture analysis method, GLCM, allows to present surface details of a sample with a mathematical function (procedural texture), and to extract the following texture descriptors: angular second moment, entropy, correlation, contrast, and inverse different moment [ 99 , 103 ]. A set of texture descriptors can be then applied as a base for a high efficiency classification tool that will allow to recognise elements of an SEM image on the basis of the texture of image pixel.…”

“…So far, there are only few literature studies reporting the application of textural analysis in biological sciences, including the detection of polysaccharide in raspberry powders [ 99 ], and cervical precancerous [ 102 ]. Nevertheless, the presence of different textures in SEM images of B35 cells, associated with a cell’s body, an extracellular matrix, neurites, etc.…”