Cell Detection Using Extremal Regions in a Semisupervised Learning Framework

Figure 3

Visual representation of cell boundaries for the different datasets (top-bottom: phase-contrast hela cells, synthetic fluorescence cells, Drosophila Kc167 cells, and fission yeast cells). The cell boundaries are shown in green, the predicted center is in red, and the true centroids are in blue. Left-right: input image, results—Arteta et al. [1], results—semisupervised framework (SS), and results—semisupervised framework with pixel classifier (SS + ILASTIK). The unlabeled data has contributions from both the training and testing images.