Figure 3From: An active learning based classification strategy for the minority class problem: application to histopathology annotationExamples of Feature Types. Examples of the feature types extracted on two ROIs from a biopsy sample (a), identified by black squares. Shown are (b), (f) the original tissue image, (c), (g) a greylevel texture image (standard deviation value), (d), (h) a Haralick texture image (entropy of the co-adjacency matrix), and (e), (i) a Gabor filter feature image. The top row (b)-(e) indicates a cancerous region, while the bottom row (f)-(i) is a benign region.Back to article page