Cowpea Images Download

Background: The traditional methods for evaluating seeds are usually performed through destructive sampling followed by physical, physiological, biochemical and molecular determinations. Whilst proven to be effective, these approaches can be criticized as being destructive, time consuming, labor intensive and requiring experienced seed analysts. Thus, the objective of this study was to investigate the potential of computer vision and multispectral imaging systems supported with multivariate analysis for high-throughput classification of cowpea (Vigna unguiculata) seeds. An automated computer-vision germination system was utilized for uninterrupted monitoring of seeds during imbibition and germination to identify different categories of all individual seeds. By using spectral signatures of single cowpea seeds extracted from multispectral images, different multivariate analysis models based on linear discriminant analysis (LDA) were developed for classifying the seeds into different categories according to ageing, viability, seedling condition and speed of germination.

Conclusion: The results demonstrated the capability of the multispectral imaging system in the ultraviolet, visible and shortwave near infrared range to provide the required information necessary for the discrimination of individual cowpea seeds to different classes. Considering the short time of image acquisition and limited sample preparation, this stat-of-the art multispectral imaging method and chemometric analysis in classifying seeds could be a valuable tool for on-line classification protocols in cost-effective real-time sorting and grading processes as it provides not only morphological and physical features but also chemical information for the seeds being examined. Implementing image processing algorithms specific for seed quality assessment along with the declining cost and increasing power of computer hardware is very efficient to make the development of such computer-integrated systems more attractive in automatic inspection of seed quality.

DOWNLOAD 🔥 https://ssurll.com/2y5I2V 🔥

Cowpeas came to the Americas more recently, in the 16th or 17th centuries AD, where indigenous peoples quickly incorporated them as a staple food. In the U.S. today, a type of cowpea called black-eyed pea is both a valuable source of protein and a cultural icon.

One of the main constraints in seed production is the heterogeneity of the seed lots used during planting stage which affects growing practices and the optimum harvest time due to differential maturity of the seeds [7, 8]. Moreover, seeds undergo many forms of physiological and physicochemical alterations during storage, called ageing leading to loss in seed viability. The rate at which the seed ages depends on its ability to resist degradation and on its protective mechanisms [9]. Seed ageing is now well recognized as the major cause of reduced vigor and viability, which involves the process of deterioration and culminates in complete loss of the ability to germinate. Seed deterioration is accompanied by a cascade of physiological and biochemical perturbations resulting in reduced overall germination performance, lower speed and uniformity of germination, inferior seedling emergence and growth, reduced storability, as well as susceptibility to environmental and biological stresses, thereby resulting in a large number of abnormal seedlings and poor plant development. The low viability of seeds may also result from pest infestation or damage during drying, storage and/or any other postharvest processes. The ideal strategy for improving the overall quality in a seed lot is to screen out damaged, abnormal and non-viable seeds [6] to increase the seed uniformity and guarantee optimized plant growth protocols and yield production on farms. Segregating damaged, infected and diseased seeds from the sound ones will considerably increase the quality and the economic value of a seed batch. Thus, knowledge regarding seed vigor and viability is extremely significant for optimizing a future profitable production of cowpea.

In fact, it is a very optimistic approach to find a technology that can be implemented to determine in advance which seeds are able to produce normal plants and which ones are dead or produce abnormal plants. To the best of our knowledge this is the first study to integrate computer-vision and multispectral imaging systems in combination with chemometric multivariate analysis for non-destructive quality estimation of single/individual cowpea seeds. Thus, the main aim of this study was to employ computer vision and multispectral imaging with linear discriminant modeling to differentiate between aged and non-aged seeds and the viable and non-viable seeds and explore the possibility of discriminating germination speed of individual seeds as an indication of seed vigor.

where tag_hash_111 corresponds to a wavelength of 780 nm, and an empirically determined threshold T was used at this wavelength band. The resulting binary image was used as a mask to identify the seed pixels in the image as the main regions of interest (ROIs). This mask was applied for all bands (from 375 to 970 nm) in the multispectral image highlighting only the seeds in black background in all bands. Next, all seeds were collected in a blob database from which different attributes of the seeds such as color, dimensions, texture, shape and main spectral features of all individual seeds appeared in the image could be extracted. The extracted spectral signatures of the seeds represent the mean intensity of the reflected light at each single wavelength calculated from all seed pixels in the image. Hence, the mean reflectance spectrum of any seed in the image was represented by 20 values calculated by averaging the intensity of pixels within the ROI of this seed at the 20 bands from 375 to 970 nm. In total, 501 average spectra representing the spectral signatures of the 501 cowpea seeds involved in this study were saved and then congregated altogether in one matrix (X) to be correlated with their corresponding germination data (Y). Figure 1 shows all key steps involved in the procedure of processing multispectral images for extracting spectral information of the seeds and building the multivariate classification models.

a Main reflectance signatures of non-aged (control) and aged seeds for different periods of artificial accelerated ageing (24, 48, 72 and 96 h), b PCA score plot of the raw spectral data of all cowpea seeds showing differentiation between aged and non-aged seeds, c main reflectance signatures of germinated and non-germinated seeds despite artificial accelerated ageing implemented

Table 4 shows the performance of another LDA model built for classifying cowpea seeds into two categories: (1) seeds that are able to produce normal seedlings and (2) seeds that produced abnormal seedlings due to deformed essential structures or missing one or more of their essential structures. The results revealed that the accuracy of the model was moderate with overall classification accuracy of 68.08, 64.34 and 62.00% in training, cross-validation and validation datasets, respectively. To improve the performance of such a model, more seeds should be involved to include all possible variation in considerations.

The discrimination results between aged and non-aged seeds presented in Table 2 are in agreement with those obtained from PCA illustrated in Fig. 2b indicating that the physicochemical changes occurred in cowpea seeds during ageing were reflected in the spectral signatures of the seeds that facilitate their discrimination. In fact, during accelerated ageing lipid peroxidation leads to deterioration of cell membranes and reducing seed viability. These biochemical changes may be the reason for a clear grouping between aged and non-aged seeds when performing principle component analysis and linear discriminant analysis [13]. The degree of overlapping observed between aged seeds (AA24, AA48, AA72 and AA96) could be ascribed to the differential rates of deterioration of individual seeds, which in turn could be associated to initial health, vigor and size of these seeds. The variability in seed size within a seed lot was the main source of the variation in seed deterioration during accelerated ageing because the smaller seeds deteriorate faster than larger seeds [41]. In general, the results demonstrated that the visible and NIR spectral data extracted from multispectral images of cowpea seeds contained the required chemical information for rapid discrimination of non-aged seeds and sorting out the aged seeds. This result is of great significance in improving seed quality by excluding naturally aged seeds due to severe and unstable storage conditions and those seeds severely desiccated just before harvesting. Also, the importance of this discriminant model is also very significant in predicting the optimum conditions required for preserving seed quality during storage and for making a decision on the longest duration of seed storage without deterioration.

GE designed this study, acquired and analyzed the multispectral images, developed the model and wrote the first draft of the manuscript. M-HW prepared the seeds with different accelerated ageing, carried out the germination experiments and computer vision data extraction and contributed in writing and revision of the manuscript. DD helped in designing the experiments, developed the image analysis methods for the computer vision system and contributed in writing and revision of the manuscript. NM, JV and EB assisted in designing the experiments and in revising the manuscript. DR assisted in developing the methods, experimental design, conceived and supervised the whole experimental protocols, wrote and revised the manuscript. All authors read and approved the final manuscript.

Purdue Improved Cowpea Storage bags are opened during a May 2011 ceremony in Pala, Chad. The hermetic grain storage bags have allowed farmers in Africa to safely store their cowpea grain so they can sell their crop well beyond harvest. (Purdue Agricultural Communication photo/Beksoubo Damienne) 17dc91bb1f