Search

Lentil and field pea are each commonly marketed as split and dehulled product. For plant-breeding programmes, the genetic improvement in split-yield is a targeted trait. However, the standard laboratory method for assessment of split-yield requires milled grain to be manually sorted into split and dehulled fractions. This process is time-consuming and impacts the number of germplasm lines that can be evaluated. A machine vision approach, based on artificial neural networks, was proposed to classify split and dehulled fractions from multispectral images of grains. Three neural networks were trained on different inputs derived from the images. The networks were: (1) a convolutional network trained on the full images, (2) a convolutional network trained on distributions of image-features, and (3) a fully connected network trained on mean and standard deviation values of image-features. The accuracy and training times were compared to determine the trade-offs between training networks with smaller inputs for computational efficiency and full-image inputs for accuracy. The networks with reduced input-data dimensionality completed network training and predictions in half the time of the image-based network. The convolutional network based on the distributions of image-features achieved a validation accuracy of 88.1%. On average, this was 1.6% greater than the image-based convolutional network and 4.6% greater than the fully connected network based on simple (mean and standard deviation) features. Feature-distributions extracted from the multispectral images captured the diversity of image data required to differentiate milling categories, leading to gains in computational efficiency over the image-based network without loss of network generality.

A global predictive model was developed for protein, moisture, and grain type, using near infrared (NIR) spectra. The model is a deep convolutional neural network, trained on NIR spectral data captured from wheat, barley, field pea, and lentil whole grains. The deep learning model performs multi-task learning to simultaneously predict grain protein, moisture, and type, with a significant reduction in prediction errors compared to linear approaches (e.g., partial least squares regression). Moreover, it is shown that the convolutional network architecture learns much more efficiently than simple feedforward neural network architectures of the same size. Thus, in addition to improved accuracy, the presented deep network is very efficient to implement, both in terms of model development time, and the required computational resources.

Background and objectives Lentil (Lens Culinaris. Medik) is a highly nutritious food staple widely consumed within India subcontinent and the Mediterranean region. Although gaining popularity in western diets, wheat will continue to be a major crop as it can be used to manufacture a wide range of products. The nutritional benefits of lentils are acknowledged, particularly as a source of high protein so the incorporation of lentil flour into wheat‐based foods has the potential to improve the nutritive value of range food products. Twelve blended flours were made using different concentrations of red lentil cotyledon, wheat, and additional gluten. A blending study was undertaken to access yeast vitality, rheological properties of dough and baking characteristic of resulting bread. Findings High ratio blends of lentil flour had no negative effect on yeast vitality even at the highest concentration of 40%. Increasing substitution of lentil flour was highly correlated to increased protein (r = 0.98) and ash (r = 0.95) and a concomitant decrease in dough strength but not extensibility. Loaf volume and baking quality were also compromised at higher concentrations. At a concentration of 5% lentil flour, there were no deleterious effects on dough quality traits or on baking quality. The addition of bakers’ gluten 0.1 g/gram flour had a restorative effect on the rheological and baking characteristics of wheat–lentil composites at higher concentrations of up to 20%. Conclusions Our results show that optimal baking quality of wheat–lentil flour can be achieved using either low concentrations of up to 5% lentil flour or up to 20% lentil flour with the addition of gluten which maintained a superior loaf and crumb quality. Significance and novelty The protein and ash content of baked breads significantly increased when wheat was partially substituted with lentil flour. The concentration of lentil flour decreased dough strength and dough development time and decreased loaf volume while increasing crumb firmness in resulting bread. The addition of gluten improved the rheological and product quality of bread which allowed higher concentrations of lentil flour to be used in bread making. Balancing the ratio of lentil flour and gluten to optimize the rheological properties will result in a composite wheat–lentil bread with acceptable baking performance and enhanced nutritional benefits for consumers.

Background and objectives Lentil (Lens culinaris M.) is a high-value grain used traditionally as a minimally processed product. Lentil crops are well-suited to Mediterranean-type climates with mild winters and hot dry summers which results in the production of high-value grain. However, extreme weather conditions, such as frost, can impact the lentil seed, for example, a darkened seed color and distorted seed shape resulting in a downgrade of its market value. The quality parameters of lentil when milled as wholemeal flour is not reliant on visual or physical seed traits such as seed size and color which impact on the market value in traditional market specifications. Instead, quality parameters such as those applied in the assessment of wheat, these include flour yield and color would be appropriate parameters to determine the value of lentil flour. It is proposed that flour, even from down-graded lentil could be supplemented with wheat flour and used to enhance the quality profile of baked products including breads, pastas, and cookies. Findings This study investigated the use of premium and frost-damaged lentil flour in cookie making. Overall cookies made from wheat–lentil composite blends resulted in flatter and wider cookies that were darker than the cookies made from 100% wheat flour. Cookies made by incorporating lentil and wheat flour resulted in a significant increase in total protein, insoluble fiber, and oligosaccharides (p < .05). In addition, the phenolic acids kaempferol and procyanidin were detected in cookies made using wheat–lentil contributing to an increase in antioxidant activity (p < .05). These phenolic acids were not detected in cookies made from 100% wheat flour. Our results show that using concentrations of up to 25% lentil flour resulted in cookies that had expectable hardness and color characteristics in comparison with a 100% wheat biscuit. Conclusion This research showed that the nutritional value and functional properties of cookies can be significantly enhanced by using either premium or down-graded lentil and the quality of the cookies was not impacted by the market-grade of the raw material. Significance and novelty Investigations into the use of pulse flours such as lentil in novel food products including pasta and snack foods are gaining popularity particularly by the health-conscious consumer. The visual appearance of lentil is not critical when utilized in such products. Lentil that has been visually or physically affected by adverse weather conditions, but its nutritional and functional values have not been compromised may prove a viable inclusion in the creation of highly nutritious food alternatives. The use of down-graded lentil as flour in novel food production could ultimately add value to pulse grain affected by adverse environmental conditions.

Post-harvest change in the colour of green field pea (Pisum sativum L.) is undesirable as this impacts the visual quality and market value of the seed. To date, there is no standard, objective method to determine bleaching. Therefore, the aim of this study was to develop an objective method for scoring bleaching based on colour reflectance spectra, measured both by spectrophotometer and multispectral Image Analysis (IA). Green field pea seeds were sorted into samples of uniform colour and these were used to train the model. Spectra calculated from multispectral images (with colour bands at 405,470,530,590,660 and 850nm) were matched to the spectrophotometer output through multiple linear regression. All spectra were transformed to emphasize the wavelength regions most impacted during bleaching, following which two critical reflectance values were scaled to a single bleaching score. The bleaching assessment method was tested in a time-course experiment comprising seeds from five green-pea genotypes stored for six months. Each sample was divided into two so that half of the seeds were stored in the dark and the remainder were exposed to controlled light to exaggerate bleaching. Throughout this period, the samples were imaged at six-weekly intervals. Assessment of bleaching by the IA method agreed well with spectrophotometer measurements, achieving a Lin's concordance statistic of 0.99 and 0.96 for the calibration and time-course samples respectively. The IA method proved more versatile because assessments could be made on individual seeds enabling the computation of bleaching uniformity within each sample. This method captured differences between genotypes in the extent, rate and uniformity of bleaching. All genotypes exhibited susceptibility to bleaching when stored under the controlled light conditions. Excell was observed to be the most susceptible genotype with the greatest bleaching-rate and OZB1308 displayed the most colour-stability.

Field peas (Pisum sativum L.) are generally traded based on seed appearance, which subjectively defines broad market-grades. In this study, we developed an objective Linear Discriminant Analysis (LDA) model to classify market grades of field peas based on seed colour, shape and size traits extracted from digital images. Seeds were imaged in a high-throughput system consisting of a camera and laser positioned over a conveyor belt. Six colour intensity digital images were captured (under 405, 470, 530, 590, 660 and 850nm light) for each seed, and surface height was measured at each pixel by laser. Colour, shape and size traits were compiled across all seed in each sample to determine the median trait values. Defective and non-defective seed samples were used to calibrate and validate the model. Colour components were sufficient to correctly classify all non-defective seed samples into correct market grades. Defective samples required a combination of colour, shape and size traits to achieve 87% and 77% accuracy in market grade classification of calibration and validation sample-sets respectively. Following these results, we used the same colour, shape and size traits to develop an LDA model which correctly classified over 97% of all validation samples as defective or non-defective.