Use of variogram analysis to classify field peas with and without internal defects caused by weevil infestation

Abstract

In this study, we acquired 72 (training data) and 30 (independent validation) high-spatial resolution (7 by 7 pixels per mm 2 ) hyperspectral imaging data [240 spectral bands from 392 to 889 nm (spectral resolution = 2.1 nm)] from samples of field peas (Pisum sativum) with and without pea weevil (Bruchus pisorum) infestation. The reflectance data were analyzed with linear discriminant analysis (LDA) or either reflectance values only or of a combination of reflectance values and variogram parameters (derived from variogram analysis) from a single spectral band (782 nm). All examined classification models were assessed based on sensitivity (ability to positively detect infestation), specificity (ability to positively detect noninfestation), and accurate classification of 30 samples of independent validation data. Highest classification performance was obtained with a combination of reflectance values in two spectral bands (641 nm and 868 nm) and variogram parameters derived from 782 nm. This classification model was associated with a sensitivity of 94.7% and a specificity of 100%. In addition, all 30 independent validation data were accurately classified (100%). For comparison, traditional linear discriminant analyses of 108,351 reflectance profiles from individual pixels in the 72 samples or average reflectance profiles from the 72 samples classified the validation data with 84.0% and 83.3% accuracy, respectively. We are unaware of any reflectance-based classification system, which -based on reflectance data acquired in only three channels (spectral bands) -can provide this level of sensitivity and specificity and classification of independent validation data of internal defects in food products. Accurate and reliable classification of food objects based reflectance values in only a few spectral bands would likely imply low computer processing requirements and rapid data analysis. Thus, we believe that the current classification method may be useful for quality control systems of a wide range of food products.

Using hyperspectral imaging to determine germination of native Australian plant seeds

“…Spectral binning was deployed, as it has been shown to increase classification accuracy [23,40]. Similar previously published studies [24,40], a radiometric filter was applied to exclude background, so that a pixel was only included, if the reflectance value of Acacia and Banksia seed coat at 660 nm (R660) met the following criterion: 0:050 < R660 < 0:250…”

“…Several studies have demonstrated the potential of reflectancebased spectroscopy methods in studies of plant seeds, including detection of internal infestations by weevils (Bruchus pisorum) in dry field peas (Pisum sativum) [23,24], classification of near isogenic maize lines (Zea mays) [25], ageing of cabbage seeds [26], classification of near isogenic maize lines (Z. mays) [27], differentiation between black walnut (Juglans nigra) shell and pulp [28], sorting of lettuce (Lactuca sativa) seeds [29], and viability of horticultural seeds [26,30,31]. These spectroscopy studies are based on the fundamental assumption that reflectance data acquired from the seed coat provides indicative information about the quality/ germination of the given seed.…”

Reflectance-based identification of parasitized host eggs and adult Trichogramma specimens

“…Compared with classical taxonomy (under the microscope) or molecular-based classification of minute and closely related animals, plant seeds and growing plants, a reflectance-based method may be of considerable relevance to a wide range of biological studies. There are numerous approaches to classification of hyperspectral imaging data, and only recently has the approach (a combination of variogram analysis and linear discriminant analysis) used in this study been described and successfully applied (Nansen, 2012;Nansen et al, 2010a;Nansen et al, 2009;Nansen et al, 2014). However, this analysis represents the first application of this classification method to the identification of animals.…”

“…The classification of reflectance data was based on a combination of variogram analysis (Nansen, 2012;Nansen et al, 2014) and linear discriminant analysis (Fisher, 1936). In brief, the analytical approach consists of conducting variogram analysis of reflectance values in individual spectral bands from each hyperspectral image.…”

Remote Sensing and Reflectance Profiling in Entomology

“…The approach has been used to detect damage and internal infestation in food products, including field peas (Pisum sativum) (100,158), wheat kernels (Triticum aestivum) (125,126), soy beans (Glycine max) (52), and jujubes (Ziziphus jujuba) (139,140). In addition, thermal imaging (reflectance in the 8-12 µm range) has been used to detect infestations by a stored grain beetle (Cryptolestes ferrugineus) inside wheat kernels (80) and infestations by insects in a wide range of other food products (136).…”