2D

This is a classical workflow for spot detection or blob like structures (vesicules, melanosomes,...)

Step 1 Laplacian of Gaussian to enhance spots . Paraeters= radius, about the average spot radius

Step 2 Detect minima (using Find Maxima with light background option to get minima). Parameter : Tolerance to Noise: to be tested, hard to predict. About the height of the enhanced feautures peaks

The best way to start writing an ImageJ2 plugin (ImageJ2 developers call it command and not plugin) is to download the example command from github and modify it. There is a video tutorial on the whole workflow on how to do this on youtube.

Scikit-learn (sklearn) is a python library used for machine learning. sklearn contains simple and efficient tools for data mining and data analysis. Modules and functions include those for classification, regression, clustering, dimensionality reduction, model selection and data preprocessing. Many people have contributed to sklearn (list of authors)

AutoPilot is the open source project that hosts the general algorithm for fast and robust assessment of local image quality, an automated computational method for image-based mapping of the three-dimensional light-sheet geometry inside a fluorescently labeled biological specimen, and a general algorithm for data-driven optimization of the system state of light-sheet microscopes capable of multi-color imaging with multiple illumination and detection arms.

The interactive Watershed Fiji plugin provides an interactive way to explore local maxima and threshold values while a resulting label map is updated on the fly.

After the user has found a reliable parameter configuration, it is possible to apply the same parameters to other images in a headless mode, for example via ImageJ macro scripting.