This interdisciplinary PhD project aimed to accelerate several data analysis tasks in ecological research that are too timeconsuming to perform manually. For some specific tasks, such as instance segmentation of tree rings or root turnover estimation, the performance of existing, generic methods failed to meet the desired level of accuracy. This led to the development of new algorithms which constitute the main part of this thesis. Incorporating domain-specific prior knowledge has shown to be essential in their development.<eng>
