Stomata (singular, “stoma”) are tiny pores through which plants breathe. Stomata are found on the upper and lower sides of leaves, on flower petals, on stems, and on roots. Scientists survey plant surfaces to determine the density and size of stomata and relate these findings to properties of the environment, such as temperature and the amounts of sunlight, humidity, oxygen, and carbon dioxide in the air when a leaf is formed. Stomata of various plants are suitable subjects for classroom laboratory activities since they may be examined by light microscopy.

Deep neural networks have attracted considerable attention because of their state-of-the-art performance on a variety of image restoration tasks, including image completion, denoising, and segmentation. However, their record of performance is built upon extremely large datasets. In many cases (for example, electron microscopy), it is extremely labor intensive, if not impossible, to acquire tens of thousands of images for a single project. The present work shows the possibility of attaining high-accuracy image segmentation, isolating regions of interest, for small datasets of transmission electron micrographs by employing encoder-decoder neural networks and image augmentation.

Microbial-induced calcite precipitation (MICP) has gained much attention in soil improvement studies, where it can enhance the physical properties of sandy soil. Small-scale sand cylinder tests were conducted to investigate the formation and failure of calcium carbonate precipitation bonding between individual sand particles. Bonding formation by precipitation was examined by scanning electron microscopy. Energy-dispersive X-ray spectroscopy analysis verified the existence of calcium, carbon, and oxygen, which could form CaCO3 after MICP-treatment. Focused-ion-beam milling was applied to study the interior structure of calcium carbonate precipitation during the MICP process.