Empirical and theoretical evidence for why Figure of Merit (FOM) is the very best edge-detection evaluation metric
Image segmentation and edge detection are closely related tasks. Take this output from a coastal segmentation model for instance:
The model will classify every pixel as either land or ocean (segmentation mask). Then the coastline is the pixels where this classification changes (edge map). On the whole, edge detection could be done using the boundaries of the output of a picture segmentation model.
I wanted to make use of this relationship in my research to assist evaluate coastal image segmentation models. Similar research all use confusion matrix-based metrics like accuracy, precision and recall. These compare all pixels in a predicted segmentation mask to a ground truth mask.
The issue is these might overestimate performance in crucial region — the coastline.
Nearly all of pixels are in the course of the ocean or completely surrounded by land. This makes them easier to categorise than those near the coastline. You’ll be able to see this in Figure 2. Unfortunately, these errors could also be shrouded within the sea of accurately classified pixels.
