Drying kinetics and the moisture distribution map of radish slabs under different drying methods (hot-air drying (HAD), microwave drying (MD), and hot-air and microwave combination drying (HMCD)) were determined and visualized by hyperspectral image (HSI) processing coupled with a partial least square regression (PLSR)-variable importance in projection (VIP) model, respectively. Page model was the most suitable in describing the experimental moisture loss data of radish slabs regardless of the drying method. Dielectric properties (DP, $\epsilon$ ) of radish slices decreased with the decrease in moisture content (MC) during MD, and the penetration depth of microwaves in radish was between 0.81 and 1.15 cm. The PLSR-VIP model developed with 38 optimal variables could result in the high prediction accuracies for both the calibration ( $R_{cal}^2=0.967$ and $\mathrm{RMSEC}=4.32\%$ ) and validation ( $R_{val}^2=0.962$ and $\mathrm{RMSEC}=4.45\%$ ). In visualized drying patterns, the radish slabs dried by HAD had a higher moisture content at the center than at the edges; however, the samples dried by MD contained higher moisture content at the edges. The nearly uniform drying pattern of radish slabs under HMCD was observed in hyperspectral images. Drying uniformity of radish slabs could be improved by the combination drying method, which significantly reduces drying time.

The feasibility of a color machine vision technique with the one-class classification method was investigated for the quality assessment of tomato seeds. The health of seeds is an important quality factor that affects their germination rate, which may be affected by seed contamination. Hence, segregation of healthy seeds from diseased and infected seeds, along with foreign materials and broken seeds, is important to improve the final yield. In this study, a custom-built machine vision system containing a color camera with a white light emitting diode (LED) light source was adopted for image acquisition. The one-class classification method was used to identify healthy seeds after extracting the features of the samples. A significant difference was observed between the features of healthy and infected seeds, and foreign materials, implying a certain threshold. The results indicated that tomato seeds can be classified with an accuracy exceeding 97%. The infected tomato seeds indicated a lower germination rate (<10%) compared to healthy seeds, as confirmed by the organic growing media germination test. Thus, identification through image analysis and rapid measurement were observed as useful in discriminating between the quality of tomato seeds in real time.

Raman imaging has been proven to be a powerful analytical technique for the characterization and visualization of chemical components in a range of products, particularly in the food and pharmaceutical industries. The conventional backscattering Raman imaging technique for the spatial analysis of a deep layer suffers from the presence of intense fluorescent and Raman signals originating from the surface layer which mask the weaker subsurface signals. Here, we demonstrated the application of a new reflection amplifying method using a background mirror as a sample holder to increase the Raman signals from a deep layer. The approach is conceptually demonstrated on enhancing the Raman signals from the subsurface layer. Results show that when bilayer samples are scanned on a reflection mirror, the average signals increase 1.62 times for the intense band at 476 cm⁻¹ of starch powder, and average increases of 2.04 times (for the band at 672 cm⁻¹) for a subsurface layer of high Raman sensitive melamine powder under a 1 mm thick teflon sheet. The method was then applied successfully to detect noninvasively the presence of small polystyrene pieces buried under a 2 mm thick layer of food powder (a case of powdered food adulteration) which otherwise are inaccessible to conventional backscattering Raman imaging. In addition, the increase in the Raman signal to noise ratio when measuring samples on a mirror is an important feature in many applications where high-throughput imaging is of interest. This concept is also applicable in an analogous manner to other disciplines, such as pharmaceutical where the Raman signals from deeper zones are typically, substantially diluted due to the interference from the surface layer.

Raman imaging requires the effective extraction of chemical information from the corresponding datasets, which can be achieved by a range of analytical methods. However, since each of these methods exhibits both strengths and weaknesses, we herein directly compare univariate, bivariate, and multivariate analyses of Raman imaging data by evaluating their performance in the quantitation of two adulterants in paprika powder. Univariate and bivariate models were developed based on the spectral features of the target adulterants, whereas spectral angle mapper (SAM), adopted as a multivariate analysis method, utilized the complete dataset. The obtained results demonstrate that despite being simple and easily implementable, the univariate method affords false positive pixels in the presence of background noise. Luckily, the above problem can be easily resolved using the bivariate method, which utilizes the multiplication of two band images wherein the same adulterant shows high-intensity peaks exhibiting the least overlap with those of other sample constituents. Finally, images produced by SAM contain abundant false negative pixels of adulterants, particularly for low-concentration samples. Notably, the bivariate method affords results closely matching the theoretical adulterant content, exhibiting the advantages of using non-complex data (only two bands are utilized) and being well suited to online applications of Raman imaging in the agro-food sector.