<> "The repository administrator has not yet configured an RDF license."^^ . <> . . "Data Fusion and Bayesian Smoothing for Tracking in Fluorescence Microscopy Images"^^ . "Obtaining spatial-temporal information of virus particles in fluorescence microscopy images is a prerequisite to gain insights into viral pathogens at a microscopic level about virus replication and assembly, but also at a macroscopic level to understand virus spread and infection in tissue-like structures. To obtain spatial-temporal information of virus particles, these structures need to be detected and tracked over time in time-lapse fluorescence microscopy image data. Since accurate manual determination of the position of many particles for all time points in microscopy image data is not feasible and introduces human bias, automatic computer vision methods for particle tracking and trajectory analysis are required.\r\n\r\nIn this thesis, new methods for probabilistic particle tracking are introduced based on data fusion and Bayesian smoothing. We propose data fusion approaches to incorporate image intensity and position as well as motion information and use a Bayesian framework to exploit uncertainties introduced by image noise and integrate a priori knowledge. To exploit image intensity and position information, we consider multiple measurements for each particle and fuse them by taking into account different uncertainties. Further, we developed a novel intensity-based probabilistic fusion approach which fuses results from multiple detectors and yields a consistent estimate of multiple fused detections to improve particle detection and localization. This approach integrates detections from classical and deep learning methods as well as exploits single-scale and multi-scale detections. To improve particle tracking by incorporating temporal information, we developed a novel Bayesian smoothing approach which integrates information from past and future time points. The covariance intersection algorithm is used to fuse position information and to obtain consistent trajectory estimates. In addition, motion information based on displacements from past and future time points is used to improve correspondence finding.\r\n\r\nThe novel methods were applied and evaluated on image data consisting of state-of-the-art benchmark data sets as well as live cell fluorescence microscopy image data showing immunodeficiency virus type-1 (HIV-1) and hepatitis C virus (HCV). It turns out, that the developed novel methods yield competitive or improved results compared to existing methods. We also applied the methods to quantify the motion and colocalization of HIV-1, HCV, and chromatin structures. Insights into viral and chromatin structures were obtained to better understand virus replication, assembly, spread and infectivity in tissue-like structures, as well as nuclear organization in mammalian cells."^^ . "2022" . . . . . . . "Christian"^^ . "Ritter"^^ . "Christian Ritter"^^ . . . . . . "Data Fusion and Bayesian Smoothing for Tracking in Fluorescence Microscopy Images (PDF)"^^ . . . "PhD_Thesis_ChristianRitter.pdf"^^ . . . "Data Fusion and Bayesian Smoothing for Tracking in Fluorescence Microscopy Images (Other)"^^ . . . . . . "indexcodes.txt"^^ . . . "Data Fusion and Bayesian Smoothing for Tracking in Fluorescence Microscopy Images (Other)"^^ . . . . . . "lightbox.jpg"^^ . . . "Data Fusion and Bayesian Smoothing for Tracking in Fluorescence Microscopy Images (Other)"^^ . . . . . . "preview.jpg"^^ . . . "Data Fusion and Bayesian Smoothing for Tracking in Fluorescence Microscopy Images (Other)"^^ . . . . . . "medium.jpg"^^ . . . "Data Fusion and Bayesian Smoothing for Tracking in Fluorescence Microscopy Images (Other)"^^ . . . . . . "small.jpg"^^ . . "HTML Summary of #32273 \n\nData Fusion and Bayesian Smoothing for Tracking in Fluorescence Microscopy Images\n\n" . "text/html" . . . "004 Informatik"@de . "004 Data processing Computer science"@en . . . "620 Ingenieurwissenschaften"@de . "620 Engineering and allied operations"@en . .