Chained Deep Learning Using Generalized Cross-Entropy for Multiple Annotators Classification.

Supervised learning requires the accurate labeling of instances, usually provided by an expert. Crowdsourcing platforms offer a practical and cost-effective alternative for large datasets when individual annotation is impractical. In addition, these platforms gather labels from multiple labelers. Still, traditional multiple-annotator methods must account for the varying levels of expertise and the noise introduced by unreliable outputs, resulting in decreased performance. In addition, they assume a homogeneous behavior of the labelers across the input feature space, and independence constraints are imposed on outputs. We propose a Generalized Cross-Entropy-based framework using Chained Deep Learning (GCECDL) to code each annotator's non-stationary patterns regarding the input space while preserving the inter-dependencies among experts through a chained deep learning approach. Experimental results devoted to multiple-annotator classification tasks on several well-known datasets demonstrate that our GCECDL can achieve robust predictive properties, outperforming state-of-the-art algorithms by combining the power of deep learning with a noise-robust loss function to deal with noisy labels. Moreover, network self-regularization is achieved by estimating each labeler's reliability within the chained approach. Lastly, visual inspection and relevance analysis experiments are conducted to reveal the non-stationary coding of our method. In a nutshell, GCEDL weights reliable labelers as a function of each input sample and achieves suitable discrimination performance with preserved interpretability regarding each annotator's trustworthiness estimation.

Automatic wheat ear counting using machine learning based on RGB UAV imagery.

In wheat (Triticum aestivum L) and other cereals, the number of ears per unit area is one of the main yield-determining components. An automatic evaluation of this parameter may contribute to the advance of wheat phenotyping and monitoring. There is no standard protocol for wheat ear counting in the field, and moreover it is time consuming. An automatic ear-counting system is proposed using machine learning techniques based on RGB (red, green, blue) images acquired from an unmanned aerial vehicle (UAV). Evaluation was performed on a set of 12 winter wheat cultivars with three nitrogen treatments during the 2017-2018 crop season. The automatic system uses a frequency filter, segmentation and feature extraction, with different classification techniques, to discriminate wheat ears in micro-plot images. The relationship between the image-based manual counting and the algorithm counting exhibited high levels of accuracy and efficiency. In addition, manual ear counting was conducted in the field for secondary validation. The correlations between the automatic and the manual in-situ ear counting with grain yield were also compared. Correlations between the automatic ear counting and grain yield were stronger than those between manual in-situ counting and GY, particularly for the lower nitrogen treatment. Methodological requirements and limitations are discussed.

Leaf dorsoventrality as a paramount factor determining spectral performance in field-grown wheat under contrasting water regimes.

The effects of leaf dorsoventrality and its interaction with environmentally induced changes in the leaf spectral response are still poorly understood, particularly for isobilateral leaves. We investigated the spectral performance of 24 genotypes of field-grown durum wheat at two locations under both rainfed and irrigated conditions. Flag leaf reflectance spectra in the VIS-NIR-SWIR (visible-near-infrared-short-wave infrared) regions were recorded in the adaxial and abaxial leaf sides and at the canopy level, while traits providing information on water status and grain yield were evaluated. Moreover, leaf anatomical parameters were measured in a subset of five genotypes. The spectral traits studied were more affected by the leaf side than by the water regime. Leaf dorsoventral differences suggested higher accessory pigment content in the abaxial leaf side, while water regime differences were related to increased chlorophyll, nitrogen, and water contents in the leaves in the irrigated treatment. These variations were associated with anatomical changes. Additionally, leaf dorsoventral differences were less in the rainfed treatment, suggesting the existence of leaf-side-specific responses at the anatomical and biochemical level. Finally, the accuracy in yield prediction was enhanced when abaxial leaf spectra were employed. We concluded that the importance of dorsoventrality in spectral traits is paramount, even in isobilateral leaves.

Cereal Crop Ear Counting in Field Conditions Using Zenithal RGB Images.

Ear density, or the number of ears per square meter (ears/m2), is a central focus in many cereal crop breeding programs, such as wheat and barley, representing an important agronomic yield component for estimating grain yield. Therefore, a quick, efficient, and standardized technique for assessing ear density would aid in improving agricultural management, providing improvements in preharvest yield predictions, or could even be used as a tool for crop breeding when it has been defined as a trait of importance. Not only are the current techniques for manual ear density assessments laborious and time-consuming, but they are also without any official standardized protocol, whether by linear meter, area quadrant, or an extrapolation based on plant ear density and plant counts postharvest. An automatic ear counting algorithm is presented in detail for estimating ear density with only sunlight illumination in field conditions based on zenithal (nadir) natural color (red, green, and blue [RGB]) digital images, allowing for high-throughput standardized measurements. Different field trials of durum wheat and barley distributed geographically across Spain during the 2014/2015 and 2015/2016 crop seasons in irrigated and rainfed trials were used to provide representative results. The three-phase protocol includes crop growth stage and field condition planning, image capture guidelines, and a computer algorithm of three steps: (i) a Laplacian frequency filter to remove low- and high-frequency artifacts, (ii) a median filter to reduce high noise, and (iii) segmentation and counting using local maxima peaks for the final count. Minor adjustments to the algorithm code must be made corresponding to the camera resolution, focal length, and distance between the camera and the crop canopy. The results demonstrate a high success rate (higher than 90%) and R2 values (of 0.62-0.75) between the algorithm counts and the manual image-based ear counts for both durum wheat and barley.

Wheat ear counting in-field conditions: high throughput and low-cost approach using RGB images.

BACKGROUND: The number of ears per unit ground area (ear density) is one of the main agronomic yield components in determining grain yield in wheat. A fast evaluation of this attribute may contribute to monitoring the efficiency of crop management practices, to an early prediction of grain yield or as a phenotyping trait in breeding programs. Currently the number of ears is counted manually, which is time consuming. Moreover, there is no single standardized protocol for counting the ears. An automatic ear-counting algorithm is proposed to estimate ear density under field conditions based on zenithal color digital images taken from above the crop in natural light conditions. Field trials were carried out at two sites in Spain during the 2014/2015 crop season on a set of 24 varieties of durum wheat with two growing conditions per site. The algorithm for counting uses three steps: (1) a Laplacian frequency filter chosen to remove low and high frequency elements appearing in an image, (2) a Median filter to reduce high noise still present around the ears and (3) segmentation using Find Maxima to segment local peaks and determine the ear count within the image. RESULTS: The results demonstrate high success rate (higher than 90%) between the algorithm counts and the manual (image-based) ear counts, and precision, with a low standard deviation (around 5%). The relationships between algorithm ear counts and grain yield was also significant and greater than the correlation with manual (field-based) ear counts. In this approach, results demonstrate that automatic ear counting performed on data captured around anthesis correlated better with grain yield than with images captured at later stages when the low performance of ear counting at late grain filling stages was associated with the loss of contrast between canopy and ears. CONCLUSIONS: Developing robust, low-cost and efficient field methods to assess wheat ear density, as a major agronomic component of yield, is highly relevant for phenotyping efforts towards increases in grain yield. Although the phenological stage of measurements is important, the robust image analysis algorithm presented here appears to be amenable from aerial or other automated platforms.

Comparative UAV and Field Phenotyping to Assess Yield and Nitrogen Use Efficiency in Hybrid and Conventional Barley.

With the commercialization and increasing availability of Unmanned Aerial Vehicles (UAVs) multiple rotor copters have expanded rapidly in plant phenotyping studies with their ability to provide clear, high resolution images. As such, the traditional bottleneck of plant phenotyping has shifted from data collection to data processing. Fortunately, the necessarily controlled and repetitive design of plant phenotyping allows for the development of semi-automatic computer processing tools that may sufficiently reduce the time spent in data extraction. Here we present a comparison of UAV and field based high throughput plant phenotyping (HTPP) using the free, open-source image analysis software FIJI (Fiji is just ImageJ) using RGB (conventional digital cameras), multispectral and thermal aerial imagery in combination with a matching suite of ground sensors in a study of two hybrids and one conventional barely variety with ten different nitrogen treatments, combining different fertilization levels and application schedules. A detailed correlation network for physiological traits and exploration of the data comparing between treatments and varieties provided insights into crop performance under different management scenarios. Multivariate regression models explained 77.8, 71.6, and 82.7% of the variance in yield from aerial, ground, and combined data sets, respectively.