Hi everyone! I am currently doing research on bacterial filaments and was wondering what the best available option would be to define the shape of individual cells with a high priority on the trainable being able to recognize gaps between the cells. I have looked at the Gaussian blur but was just wondering if there were other methods.
Images or image subsections showing what you want to separate are usually the best way to get an accurate response. I have seen everything from phase contrast to fluorescently labeled/stained bacteria.
That said, if you are interested in enhancing edges, you might want to look into the standard deviation type filters.
Unsharp masking can be amazing, but requires some idea of the size of what you want to enhance, and usually only works well if everything you are looking for is about the same size and shape.