A carregar...
Publicação
Simulation of biologically inspired object movement for the study of object tracking algorithms
| Nome: | Descrição: | Tamanho: | Formato: | |
|---|---|---|---|---|
| 1.46 MB | Adobe PDF |
Autores
Orientador(es)
Resumo(s)
Major advances in Cell and Molecular Biology have been associated with the advances in live-cell microscopy imaging, and these studies started to rely on temporal single cell imaging. To support these efforts, available automated image analysis methods such as cell segmentation and cell tracking during a time-series analysis should be improved. One important step is the validation of such image processing methods. Ideally, the “ground truth” should be known, which is possible only by manually labelling images or by artificially produced images. To simulate such artificial images we developed a platform that can simulate biologically inspired objects, by generating bodies with different morphologies, physical movement and that can aggregate in clusters. Using this platform, we tested and compared four tracking algorithms: Simple Nearest-Neighbour (NN), NN with Morphology and two DBSCAN based ones. In this work we showed that Simple NN work for small object velocities, while the other algorithms perform better on higher velocities and when clustered. This platform can generate new benchmark images and is openly available to test other tracking algorithms. (http://griduni.uninova.pt/Clustergen/ClusterGen_v1.0.zip)
Descrição
Palavras-chave
Microscopy Imaging Synthetic Time-lapse Image Simulation Cell Tracking Cluster Tracking