Extended Data Fig. 1: A test of Weka trainable segmentation with different input data. | Nature Methods

Extended Data Fig. 1: A test of Weka trainable segmentation with different input data.

From: ERnet: a tool for the semantic segmentation and quantitative analysis of endoplasmic reticulum topology

Extended Data Fig. 1

Top left: An input image was used to train a classifier of Weka Trainable Segmentation. Top right: The tubules (cyan) and sheet (yellow) can be clearly classified after segmentation. Bottom left: a new image was applied to the trained classifier shown above. Bottom right: segmentation result of the new input data. This test was independently repeated five times with similar results. Scale bars: 5 μm.

Back to article page