ZEISS Knowledge Base

Pipeline operations layout

The following operations are part of the pipeline.

Input ROI

This operation allows to select the region of interest (ROI). ROI defines the dataset subarea that will be processed and analyzed by the pipeline.

Denoising

Set of operations performing noise reduction. Choose the more suitable method based on your sample and define the diameter.

Denoising parameters

Parameter		Description
Channels		Sets the processing and analysis target channel(s).
Method		Sets the denoising algorithm.
–	Bilateral	The Bilateral filtering can reduce the noise in an image while maintaining edges. A bilateral filtering blurs an image using both domain and range neighborhoods.
–	Discrete Gaussian	The Discrete Gaussian filter blurs an image by convolution with a discrete Gaussian kernel. This method is fast, but blurs edges.
–	Flow-driven	The curvature Flow-driven denoising filter is an anisotropic diffusion method used to reduce noise or unwanted detail in images while preserving specific image features.
–	Mean	The Mean filter blurs the image by calculating a new intensity value for each pixel. The new intensity is equal to the average of the intensity values of the pixels in the local neighborhood.
–	Median	The Median filter sets the intensity for each pixel to the median of the intensity values in the local neighborhood. The median is the intensity value and the center of the ordered sequence of all pixels in the local neighborhood.
–	Particle enhancement	The Particle Enhancement Filter can be used to extract bright structures of a certain size from a noisy background. It convolves a given image with a special restoration kernel.
Diameter		Sets the reference objects the reference objects diameter. Note: The filter size is expressed as the smaller objects' diameter of the structures that you want to preserve or enhance. This parameter must be expressed in metric unit. Note: You can measure the diameter directly from the dataset with the Measurement tool.

Seeded Region Growing

Automatic objects detection algorithm based on region growing approach. It uses seeds as starting point for the growing task.

Seeded Region Growing parameters

Seeding

Parameter		Description
Method		Sets the seeding method.
–	Nucleus-based	Detects the bright peaks such as nuclei.
–	Membrane-based	Detects dark peaks inside membrane staining.
Channel		Sets the analysis target channel.
Seeded Detection		Sets the seed threshold percentage. A lower value detects fewer, but bigger seeds. A higher value detects more, but smaller seeds.
Area		Sets the Area, Volume or Diameter seeds range. Only the segments that matches the range are accepted as seeds. The filter can be enabled/disabled. Note: You can measure the diameter directly from the dataset with the Measurement tool.
–	Diameter	Filters the unwanted seeds based on their diameter.
–	Volume	Filters the unwanted seeds based on their volume.
–	Area	Filters the unwanted seeds based on their area.

Region Growing

Parameter		Description
Method		Sets the region growing method.
–	Watershed	Segments bright objects such as cells by growing from seeds to the border of an object.
–	Membrane-based watershed	Segments dark objects such as membrane staining by growing from the seeds to the bright intensity outlines.
Channel		Sets the channel for the selected algorithm.
Threshold		Sets this parameter so that the foreground of the region growing channel contains the objects of interest.
Max. distance		Sets the maximum distance on which the region growing will act. Note: You can measure the distance directly from the dataset with the Measurement tool.

Store Objects

Store the detected segments (tag) in the active dataset.