Super resolution for root imaging.

PREMISE: High-resolution cameras are very helpful for plant phenotyping as their images enable tasks such as target vs. background discrimination and the measurement and analysis of fine above-ground plant attributes. However, the acquisition of high-resolution images of plant roots is more challenging than above-ground data collection. An effective super-resolution (SR) algorithm is therefore needed for overcoming the resolution limitations of sensors, reducing storage space requirements, and boosting the performance of subsequent analyses. METHODS: We propose an SR framework for enhancing images of plant roots using convolutional neural networks. We compare three alternatives for training the SR model: (i) training with non-plant-root images, (ii) training with plant-root images, and (iii) pretraining the model with non-plant-root images and fine-tuning with plant-root images. The architectures of the SR models were based on two state-of-the-art deep learning approaches: a fast SR convolutional neural network and an SR generative adversarial network. RESULTS: In our experiments, we observed that the SR models improved the quality of low-resolution images of plant roots in an unseen data set in terms of the signal-to-noise ratio. We used a collection of publicly available data sets to demonstrate that the SR models outperform the basic bicubic interpolation, even when trained with non-root data sets. DISCUSSION: The incorporation of a deep learning-based SR model in the imaging process enhances the quality of low-resolution images of plant roots. We demonstrate that SR preprocessing boosts the performance of a machine learning system trained to separate plant roots from their background. Our segmentation experiments also show that high performance on this task can be achieved independently of the signal-to-noise ratio. We therefore conclude that the quality of the image enhancement depends on the desired application.

Using Atmospheric Dispersion Theory to Inform the Design of a Short-lived Radioactive Particle Release Experiment.

Atmospheric dispersion theory can be used to predict ground deposition of particulates downwind of a radionuclide release. This paper uses standard formulations found in Gaussian plume models to inform the design of an experimental release of short-lived radioactive particles into the atmosphere. Specifically, a source depletion algorithm is used to determine the optimum particle size and release height that maximizes the near-field deposition while minimizing both the required source activity and the fraction of activity lost to long-distance transport. The purpose of the release is to provide a realistic deposition pattern that might be observed downwind of a small-scale vent from an underground nuclear explosion. The deposition field will be used, in part, to study several techniques of gamma radiation survey and spectrometry that could be used by an On-Site Inspection team investigating such an event.

PRex: An Experiment to Investigate Detection of Near-field Particulate Deposition from a Simulated Underground Nuclear Weapons Test Vent.

A radioactive particulate release experiment to produce a near-field ground deposition representative of small-scale venting from an underground nuclear test was conducted to gather data in support of treaty capability development activities. For this experiment, a CO2-driven "air cannon" was used to inject (140)La, a radioisotope of lanthanum with 1.7-d half-life and strong gamma-ray emissions, into the lowest levels of the atmosphere at ambient temperatures. Witness plates and air samplers were laid out in an irregular grid covering the area where the plume was anticipated to deposit based on climatological wind records. This experiment was performed at the Nevada National Security Site, where existing infrastructure, radiological procedures, and support personnel facilitated planning and execution of the work. A vehicle-mounted NaI(Tl) spectrometer and a polyvinyl toluene-based backpack instrument were used to survey the deposited plume. Hand-held instruments, including NaI(Tl) and lanthanum bromide scintillators and high purity germanium spectrometers, were used to take in situ measurements. Additionally, three soil sampling techniques were investigated and compared. The relative sensitivity and utility of sampling and survey methods are discussed in the context of on-site inspection.