Potential of Artificial Neural Networks in Varietal Identification using Morphometry of Wheat Grains

Dubey, B. P.; Bhagwat, S. G.; Shouche, S. P.; Sainis, Jayashree Krishna

doi:10.1016/j.biosystemseng.2006.06.001

Cited by 91 publications

(45 citation statements)

References 13 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…According to Choudhary et al (2008), the classification of wheat grain samples with an accuracy of 89.4% for Canada Western red spring (CWRS) wheat and 99.3% for Canada Western amber durum (CWAD) wheat was possible based on combining the morphological, colour, textural and wavelet features. Dubey et al (2006), using 45 morphometric features and ANN, correctly identified from 84% to 94% of the analysed wheat cultivars. The possibility of wheat cultivar discrimination / classification is the result of the inheritance of genes related to dimension and shape (Bergman et al, 2000;Okamoto et al, 2012), colour (Himi, Noda, 2005;Ficco et al, 2014) and endosperm hardness (Morris, Beecher, 2012), which accounts for the inter-cultivar differentiation.…”

mentioning

confidence: 99%

“…Data concerning differences of kernel dimensions within the cultivar samples are quite scarce. According to Dubey et al (2006), the changes of length and width of grain of three wheat cultivars differing in sowing date were up to 5.7%. A higher variation of length and width (up to 6.5% and to 12.1%, respectively) among recombinant lines of wheat in relation to location and year was noted by Ramya et al (2010).…”

mentioning

confidence: 99%

See 1 more Smart Citation

Grain morphology, texture and colour-related compounds of bread wheat cultivars in relation to cultivation regimes and growing location

Tańska

Buczek

Jarecki

et al. 2018

Zemdirbyste-Agriculture

View full text Add to dashboard Cite

Digital wheat grain classification systems are primarily based on kernel image and texture features. These traits are highly environment-dependent, so sample-to-sample variation may affect results of grain classification, grading, or discrimination. The main aim of this study was to determine the intra-cultivar variation of wheat kernel size, shape, colour, endosperm hardness and selected chemical compounds (carotenoids, phlobaphenes and proteins) associated with kernel colour. In the study, four winter bread wheat cultivars: two self-pollinated 'Bogatka' and 'Batuta' and two hybrid cultivars 'Hybred' and 'Hymack' were grown in different environments (extensive and intensive cultivation regimes, and three growing locations) to create variation in the kernel image, hardness and chemical compounds to simulate the range of variations encountered in reality. Measurements of kernel colour and dimensions were conducted using digital image analysis, hardness and protein content were analysed using nearinfrared reflectance (NIR) spectroscopy, while carotenoids and phlobaphenes were examined using colorimetric assays. The results showed that kernel dimensions and shape were highly intra-cultivar constant, with the highest variation of thickness (up to 7.59%) and volume (up to 10.10%). Surface colour was also relatively constant, with saturation as the most variable feature (up to 11.44%). In contrast, endosperm colour revealing an approx. 2-3-fold higher variation of colour attributes than surface colour, and the highest variation was observed in the case of saturation of the colour (up to 19.67%). Among the tested chemical compounds, carotenoids were the least intra-cultivar variable (up to 14.39%), while phlobaphenes the most intra-cultivar variable (up to 31.87%). Grain hardness was the most intra-cultivar variable feature of the tested grain samples and reached the highest value of 49.02% for cultivar 'Hymack'. The genotype had the greatest effect on kernel colour (up to 50.49% of the source of variation) and carotenoids (52.97%). The environment affected mostly the grain proteins (82.20%) and hardness (72.61%).

show abstract

mentioning

confidence: 99%

mentioning

confidence: 99%

Grain morphology, texture and colour-related compounds of bread wheat cultivars in relation to cultivation regimes and growing location

Tańska

Buczek

Jarecki

et al. 2018

Zemdirbyste-Agriculture

View full text Add to dashboard Cite

show abstract

“…There have been more studies on appearance characteristics of wheat kernels with machine vision at home and abroad, which have gained approving effect [1][2][3][4][5][6]. However, it was difficult to describe the wheat kernel shape accurately with traditional description methods because the views of wheat shape are too complex.…”

Section: Introductionmentioning

confidence: 99%

Research on the Shape of Wheat Kernels Based on Fourier Describer

Wei

Quan

2012

Computer and Computing Technologies in Agriculture V

View full text Add to dashboard Cite

Abstract. The shape of wheat kernels is one of the most important criterions for quality inspection and grading. This paper has used Fourier describer to describe the three views of wheat shape accurately and has had a counterconstruction operation.Also it verified this method for describing the shape, small but complex. Finally, the appropriate characteristic parameters were selected, and the BP-network was used to classify the four varieties wheat kernels. This method deserves a recognition rate of 98%~99%.

show abstract

“…Artificial Neural Networks (Visen et al, 2002;Dubey et al, 2006) application which has a significant potential for the classification and definition of agricultural products and used in biological practices frequently for classifications and product definitions in recent years was performed. Artificial Neural Networks (ANN) are parallel and distributed structures consisting of process elements developed through inspiration from human brain and linked together through weighted connections.…”

Section: Introductionmentioning

confidence: 99%

UPGMA and artificial neural networks applications on wild type olives

Sesli¹,

oğlu²,

Altıntaş³

et al. 2017

JEB

View full text Add to dashboard Cite

Aim: Methodology:Results: Interpretation:Plant genetic sources are important to study genetic variability and richness of hereditary knowledge of plant species in gene pool. Local varieties, rural populations, wild types and old varieties are the primary ones. In this respect, wild type olives (Olea europaea oleaster) are valuable in terms of olive breeding, cultivation and ecosystem. The aim of the study was to determine genetic distances between olive varieties.Artificial Neural Networks intuitive algorithm application was performed on seven wild type olives grown in different regions of Turkey by using data obtained from twenty-two ISSR primers.UPGMA dendrograms were developed through Jaccard, simple matching coefficients, and similarity matrices; and genetic similarities and dissimilarities were exhibited.It was concluded that Artificial Neural Networks would be beneficial for estimating olive types accurately based on the results obtained from earlier studies performed with genetic markers.

show abstract

Potential of Artificial Neural Networks in Varietal Identification using Morphometry of Wheat Grains

Cited by 91 publications

References 13 publications

Grain morphology, texture and colour-related compounds of bread wheat cultivars in relation to cultivation regimes and growing location

Grain morphology, texture and colour-related compounds of bread wheat cultivars in relation to cultivation regimes and growing location

Research on the Shape of Wheat Kernels Based on Fourier Describer

UPGMA and artificial neural networks applications on wild type olives

Contact Info

Product

Resources

About