Activated Carbon Derived from Pyrolyzed Pinewood Char using Elevated Temperature, KOH, H3PO4, and H2O2

This technical note determines the feasibility of using an InceptionV4_ResNetV2 convolutional neural network (CNN) to correctly identify hardwood species from macroscopic images. The method is composed of a commodity smartphone fitted with a 14× macro lens for photography. The end-grains of ten different North American hardwood species were photographed to create a dataset of 1869 images. The stratified 5-fold cross-validation machine-learning method was used, in which the number of testing samples varied from 341 to 342. Data augmentation was performed on-the-fly for each training set by rotating, zooming, and flipping images. It was found that the CNN could correctly identify hardwood species based on macroscopic images of its end-grain with an adjusted accuracy of 92.60%. With the current growing of machine-learning field, this model can then be readily deployed in a mobile application for field wood identification.

Native grasses and native wildflowers are declining, especially along roadside right-of-ways because of intensive mowing and herbicide management practices. Roadside right-of-ways undergo regular disturbances such as mowing, maintenance, and road developments that affect soils, groundwater, surface hydrology, and vegetation composition. We investigated species richness and percent coverage within plant communities along highway right-of-ways to determine if reduced mowing increased native plant coverage. The study was conducted using 10 research plots situated along Highway 25 in Oktibbeha and Winston counties, Mississippi. Each research plot consisted of three different treatments as follows: one that included greater than four mowings per year, one mowing only in fall, and one mowing only in fall with a supplemental native wildflower seeding. Using line transect sampling, we detected 277 plant species, which included native and nonnative forbs, legumes, grasses, rushes, sedges, and woody perennials (vines, shrubs, and trees). Total percent coverage of native and nonnative plants within different growth form categories did not differ among treatments (F2,96 = 0.45, P = 0.83). However, coverage differed between uplands and lowlands (F1,96 = 18.22, P ≤ 0.001), between years (F1,96 = 14.54, P ≤ 0.001), between fall and spring seasons (F1,96 = 16.25, P ≤ 0.001), and interacted between years and seasons (F1,96 = 24.08, P ≤ 0.001) and seasons and elevations (F1,96 = 5.00, P ≤ 0.001). Nonnative agronomic grasses exhibited the greatest coverage (> 90%) in all treatments. Percent coverage of each plant growth form was greatest in lowlands. Our research showed an increase of native grasses and wildflower species along roadsides with a reduced mowing regimen. We concluded that the timing and intensity of mowing for the duration of our study had little effect on the species composition of plant communities. However, one mowing per year retained agronomic plant coverage for erosion control and soil stabilization during roadside maintenance. Specific proactive management implementations can include native plantings, selective herbicide use to decrease nonnative grasses, continual mowing from roadside edge to 10 m, and only one mowing in late fall, but with an extension of the boundary to reach beyond 10 m from the roadside edge to suppress the invasion of woody plants, which could lead to lower long-term maintenance costs.

Microscopic wood identification plays a critical role in many economically important areas in wood science. Historically, producing and curating relevant and representative microscopic cross-section images of wood species is limited to highly experienced and trained anatomists. This manuscript demonstrates the feasibility of generating synthetic microscopic cross-sections of hardwood species. We leveraged a publicly available dataset of 119 hardwood species to train a style-based generative adversarial network (GAN). The proposed GAN generated anatomically accurate cross-section images with remarkable fidelity to actual data. Quantitative metrics corroborated the capacity of the generative model in capturing complex wood structure by resulting in a Fréchet inception distance score of 17.38. Image diversity was calculated using the Structural Similarity Index Measure (SSIM). The SSIM results confirmed that the GAN approach can successfully synthesize diverse images. To confirm the usefulness and realism of the GAN generated images, eight professional wood anatomists in two experience levels participated in a visual Turing test and correctly identified fake and actual images at rates of 48.3 and 43.7%, respectively, with no statistical difference when compared to random guess. The generative model can synthesize realistic, diverse, and meaningful high-resolution microscope cross-section images that are virtually indistinguishable from real images. Furthermore, the framework presented may be suitable for improving current deep learning models, helping understand potential breeding between species, and may be used as an educational tool.