multi-template matching
multi-template matching detects and localizes biological objects in 2D microscopy images to improve object-detection and classification for biomedical imaging and high-content screening.
Key Features:
- Multiple Template Utilization: Uses multiple template images to enhance detection of entire, partial, and multiple biological objects within complex images.
- Robustness to Challenging Contrast: Addresses limitations of classical image-segmentation methods such as thresholding and edge detection for poorly contrasted samples.
- 2D Image Analysis: Operates on 2D microscopy images for object localization tasks.
- High-Content Screening Compatibility: Applicable to large-scale datasets generated by automated microscopy and high-content screening.
- Time-Lapse Detection and Tracking: Supports detection and tracking of objects in time-lapse assays.
- Classification by Template Correspondence: Enables classification of detected regions by matching to templates corresponding to distinct object categories.
- Pipeline Integration: Can be incorporated as a general image-analysis step within custom processing pipelines.
Scientific Applications:
- Zebrafish and Medaka Localization: Localizes biological objects in zebrafish and medaka high-content screening datasets.
- Automated Microscopy Analyses: Applied to large-scale automated microscopy datasets for object detection and screening.
- Time-Lapse Assays: Used for detection and tracking tasks in time-lapse imaging experiments.
Methodology:
Matches multiple template images against target images (template matching), providing improved detection over single-template approaches and addressing limitations of thresholding and edge detection.
Topics
Details
- Tool Type:
- plugin
- Programming Languages:
- Python
- Added:
- 1/18/2021
- Last Updated:
- 3/2/2021
Operations
Publications
Thomas LSV, Gehrig J. Multi-template matching: a versatile tool for object-localization in microscopy images. BMC Bioinformatics. 2020;21(1). doi:10.1186/s12859-020-3363-7. PMID:32024462. PMCID:PMC7003318.