Viability Prediction of Ricinus cummunis L. Seeds Using Multispectral Imaging

Abstract

The purpose of this study was to highlight the use of multispectral imaging in seed quality testing of castor seeds. Visually, 120 seeds were divided into three classes: yellow, grey and black seeds. Thereafter, images at 19 different wavelengths ranging from 375–970 nm were captured of all the seeds. Mean intensity for each single seed was extracted from the images, and a significant difference between the three colour classes was observed, with the best separation in the near-infrared wavelengths. A specified feature (RegionMSI mean) based on normalized canonical discriminant analysis, were employed and viable seeds were distinguished from dead seeds with 92% accuracy. The same model was tested on a validation set of seeds. These seeds were divided into two groups depending on germination ability, 241 were predicted as viable and expected to germinate and 59 were predicted as dead or non-germinated seeds. This validation of the model resulted in 96% correct classification of the seeds. The results illustrate how multispectral imaging technology can be employed for prediction of viable castor seeds, based on seed coat colour.

Multispectral Imaging for Plant Food Quality Analysis and Visualization

“…In their studies, PLSDA prediction results demonstrated that larger seeds had higher viability and vigor compared with smaller seeds (Shetty and others ; Kandpal and others ). Also, the prediction of castor seed viability by using normalized canonical discriminant analysis (nCDA) resulted in a 96% classification accuracy, and it confirmed the feasibility to apply multispectral technology in seed viability testing (Olesen and others ). To classify rice and soybean seeds according to their varieties and genetic origins, LS‐SVM and BPNN models showed better performance (over 94%) than PLSDA models (Liu and others , , ).…”

Detection of live larvae in cocoons of Bathyplectes curculionis (Hymenoptera: Ichneumonidae) using visible/near‐infrared multispectral imaging

“…Conventional imaging is deemed insufficient for the analysis of samples of similar shape and colour, and may require additional information on the chemical background of the samples for higher accuracy . NIR spectroscopy can provide chemical information on a sample; however, it is a point‐based measurement and is compromised by the spatial heterogeneity present in a sample, which may reduce accuracy . Therefore, multispectral (MS) imaging, which combines the features of conventional imaging and spectroscopy, has emerged as a potential technology with high precision for insect pest management …”

“…The MS imaging system acquires images at preferred or selected wavelengths in the form of a cube, in which the x ‐ and y ‐axes indicate the spatial position and the λ axis gives the spectral information. The advantages of the MS imaging system are that it is non‐destructive and requires minimal or no prior sample preparation . A recent study on the identification of rice weevil species, Sitophilus (Coleoptera: Curculionidae), used an MS imaging approach by selecting only images of the useful wavelengths from an originally visible/NIR hyperspectral image .…”

“…20 NIR spectroscopy can provide chemical information on a sample; 21 however, it is a point-based measurement and is compromised by the spatial heterogeneity present in a sample, which may reduce accuracy. 21,23 Therefore, multispectral (MS) imaging, which combines the features of conventional imaging and spectroscopy, has emerged as a potential technology with high precision for insect pest management. 12,13 The MS imaging system acquires images at preferred or selected wavelengths in the form of a cube, in which the x-and y-axes indicate the spatial position and the axis gives the spectral information.…”

“…The advantages of the MS imaging system are that it is non-destructive and requires minimal or no prior sample preparation. 23 A recent study on the identification of rice weevil species, Sitophilus (Coleoptera: Curculionidae), used an MS imaging approach by selecting only images of the useful wavelengths from an originally visible/NIR hyperspectral image. 20 A similar approach was used to study pupal development in the tsetse fly (Glossina palpalis gambiensis) (Diptera: Glossinidae) using NIR hyperspectral imaging.…”