作者: Dominic Mai , Philipp Fischer , Thomas Blein , Jasmin Dürr , Klaus Palme
DOI: 10.1007/978-3-642-40602-7_21
关键词:
摘要: In this paper, we aim for detection and segmentation of Arabidopsis thaliana cells in volumetric image data. To end, cluster the training samples by their size aspect ratio learn a detector shape model each cluster. While yields good cell hypotheses, additionally aligning to allows better localize detections reconstruct case low quality input We show that due more accurate localization, alignment also improves performance.