Quality classification of Jatropha curcas seeds using radiographic images and machine learning

Medeiros, André Dantas de; Pinheiro, Daniel Teixeira; Xavier, Wanderson Andrade; Silva, Laércio Junio da; Dias, Denise Cunha Fernandes dos Santos

doi:10.1016/j.indcrop.2020.112162

Cited by 61 publications

(51 citation statements)

References 20 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Regarding the internal seed structure, it was revealed a direct relationship between the deteriorated tissues (lower density in the X-ray images) and the higher number of abnormal seedlings and dead seeds [26]. Our studies have also indicated that those deteriorated tissues represent dead seeds especially when they are reaching the embryonic axis (X-ray images in Class 3).…”

Section: Discussionsupporting

confidence: 54%

Multispectral and X-ray Images for Characterization of Jatropha Curcas L. Seed Quality

Bianchini

Mascarin

Silva

et al. 2020

Preprint

View full text Add to dashboard Cite

Background: Jatropha curcas is an oilseed plant with great potential for biodiesel production. In agricultural industry, the seed quality is still estimated by manual inspection, using destructive, time-consuming and subjective tests that depend on the seed analyst experience. Recent advances in machine vision combined with artificial intelligence algorithms can provide spatial and spectral information for characterization of biological images, reducing subjectivity and optimizing the analysis process.Results: We present a new method for automatic characterization of jatropha seed quality, based on multispectral imaging (MSI) combined with X-ray imaging. We propose an approach along with X-ray images in order to investigate internal problems such as damages in the embryonic axis and endosperm, considering the fact that seed surface profiles can be negatively affected, but without reaching important internal regions of the seeds. Our studies included the application of a normalized canonical discriminant analyses (nCDA) algorithm as a supervised transformation building method to classify spatial and spectral patters according to the classes of seed quality. Spectral reflectance signatures in a range of 780 to 970 nm and the X-ray images can efficiently predict quality traits such as normal seedlings, abnormal seedlings and dead seeds.Conclusions: MSI and X-ray images have a strong relationship with physiological performance of Jatropha curcas L. These techniques can be alternative methods for rapid, efficient, sustainable and non-destructive characterization of jatropha seed quality in the future, overcoming the intrinsic subjectivity of the conventional seed quality analysis.

show abstract

Section: Discussionsupporting

confidence: 54%

Multispectral and X-ray Images for Characterization of Jatropha Curcas L. Seed Quality

Bianchini

Mascarin

Silva

et al. 2020

Preprint

View full text Add to dashboard Cite

show abstract

“…But, dead seeds occurred mainly when deteriorated tissues were reaching the embryonic axis (X-ray images in class 3). Lower grayscale in the Xray images are strongly related to lower physical integrity and less stored reserves, including protein, carbohydrates and fats [29]. Although high-vigor seeds (Lot 2) had the lowest re ectance intensity, there was a different trend in intermediate (Lot 3) and low-vigor seeds (Lot 1) because seeds with intermediate vigor had the highest re ectance mean (Fig.…”

Section: Discussionmentioning

confidence: 97%

Multispectral and X-ray images for characterization of Jatropha curcas L. seed quality

Bianchini

Mascarin

Silva

et al. 2020

Preprint

View full text Add to dashboard Cite

Background: The use of non-destructive methods with less human interference is of great interest in agricultural industry and crop breeding. Modern imaging technologies enable the automatic visualization of multi-parameter for characterization of biological samples, reducing subjectivity and optimizing the analysis process. Furthermore, the combination of two or more imaging techniques has contributed to discovering new physicochemical tools and interpreting datasets in real time.Results: We present a new method for automatic characterization of seed quality based on the combination of multispectral and X-ray imaging technologies. We proposed an approach using X-ray images to investigate internal tissues because seed surface profile can be negatively affected, but without reaching important internal regions of seeds. An oilseed plant (Jatropha curcas) was used as a model species, which also serve as a multi-purposed crop of economic importance worldwide. Our studies included the application of a normalized canonical discriminant analyses (nCDA) algorithm as a supervised transformation building method to obtain spatial and spectral patterns on different seedlots. We developed classification models using reflectance data and X-ray classes based on linear discriminant analysis (LDA). The classification models, individually or combined, showed high accuracy (>0.96) using reflectance at 940 nm and X-ray data to predict quality traits such as normal seedlings, abnormal seedlings and dead seeds. Conclusions: Multispectral and X-ray imaging have a strong relationship with seed physiological performance. Reflectance at 940 nm and X-ray data can efficiently predict seed quality attributes. These techniques can be alternative methods for rapid, efficient, sustainable and non-destructive characterization of seed quality in the future, overcoming the intrinsic subjectivity of the conventional seed quality analysis.

show abstract

“…24 A simulated scatter plot and Latent Dirichlet allocation models could possibly train algorithms to identify biologically important traits such as flowering in response to yielding with a reduced cost of time. 25,26 Morphological traits such as plant height, number of pods, pod length and number of seeds were influenced by the yielding percentage in horse gram. 27,28 The seeds and fruits of different species vary greatly in aspects of their appearance like shape, size, orientation, and structure of the embryo in relation to storage tissues.…”

Section: Discussionmentioning

confidence: 99%

Morphological traits of drought tolerant horse gram germplasm: classification through machine learning

et al. 2020

View full text Add to dashboard Cite

BACKGROUND Horse gram (Macrotyloma uniflorum (Lam.) Verdc.) is an underutilized pulse crop with good drought resistance traits. It is a rich source of protein. Conventional breeding methods for high yielding and abiotic stress tolerant germplasm are hampered by the scarcity of morphological data sets. Thus, horse gram cultivars considered for this study is classified based on prevailing growth factors showing homogenous genotype in various agro ecological zones. Nowadays, several machine learning (ML) methods are used in the field of plant phenotyping. RESULTS We adopted unsupervised learning techniques from the K‐means clustering algorithm to analyze important morphological traits: plant shoot length, total plant height, flowering percentage, number of pods per plant, pod length, number of seeds per plant, and seed length variants between germplasm. Unsupervised clustering revealed that 20 germplasm accessions were grouped in four clusters in which high‐yielding traits were predominantly observed in cluster 2. CONCLUSION These findings could guide ML‐based classification to characterize suitable germplasms on the basis of high‐yielding varieties for different agro‐ecological zones. © 2020 Society of Chemical Industry

show abstract

Quality classification of Jatropha curcas seeds using radiographic images and machine learning

Cited by 61 publications

References 20 publications

Multispectral and X-ray Images for Characterization of Jatropha Curcas L. Seed Quality

Multispectral and X-ray Images for Characterization of Jatropha Curcas L. Seed Quality

Multispectral and X-ray images for characterization of Jatropha curcas L. seed quality

Morphological traits of drought tolerant horse gram germplasm: classification through machine learning

Contact Info

Product

Resources

About