Feature Recognition Software for Minirhizotron Image Processing

该奖项是为了开发微型电子管中的特征识别软件，微型电子管是用于观察活根的管。 细根周转是生态系统碳和养分循环的重要组成部分，但方法上的挑战限制了我们在自然条件下测量根系周转的能力。 最近，使用微型电子显微镜和微型摄像设备已成为研究细根动态在生态和农艺设置的技术选择。 然而，处理微型电子管图像所需的时间过多。在这个项目中，我们将开发开源的特征识别软件，以帮助捕获和分析微型电子管图像。通过改进用户界面和整合用于识别医学图像中血管的算法，我们将自动化微型电子管数据收集的许多方面。 植物根系的产生和死亡代表了生态系统中碳和营养物质的重要通量。因此，测量根系周转率对于预测生态系统对全球变化的反应至关重要。这项软件开发研究将有助于测量根系生产和死亡率，并有助于培养跨学科研究领域的学生。

This award is for the development of software for feature recognition in minirhizotrons, which are tubes for the observation of living roots. Fine root turnover is a significant component of ecosystem carbon and nutrient cycles, but methodological challenges have limited our ability to measure root turnover under natural conditions. Recently, the use of minirhizotrons and miniaturized camera equipment has emerged as the technique of choice for studying fine root dynamics in ecological and agronomic settings. However, the time required to process minirhizotron images is excessive. In this project, we will develop open-source, feature-recognition software to assist in the capture and analysis of minirhizotron images. By improving user interfaces and incorporating algorithms used to identify blood vessels in medical images, we will automate many aspects of minirhizotron data collection. The production and mortality of plant roots represents a significant flux of carbon and nutrients through ecosystems. Measuring rates of root turnover is therefore crucial to predicting ecosystem responses to global change. This software-development research will aid in measuring root production and mortality, and it will help train students in an interdisciplinary research area.