From: Survey statistics of automated segmentations applied to optical imaging of mammalian cells
Segmentation category | Description | Number of surveyed papers |
---|---|---|
Active contours + Level Set | Parametric curves which fit to an image object of interest. These curve fitting functions are regularized gradient edge detectors | |
Graph-based | Applies graph theories to segment regions of interest | |
Morphological | Apply morphological operations to segment or clean a pre-segmented image | |
Other | The methods in this category are created for a specific problem or cell line by a combination of existing techniques or by creating a new concept | |
Partial Derivative Equations | Groups pixels into different segment based on minimizing a cost function using partial derivatives | |
Region growing | Starts from a seed and grows the segmented regions following some pre-defined criterion | |
Thresholding | Threshold based techniques consider the foreground pixels to have intensity values higher (or lower) than a given threshold. | |
Watershed | Mainly used to separate touching cells or touching subcellular regions |