Skip to main content
Figure 4 | BMC Bioinformatics

Figure 4

From: Aro: a machine learning approach to identifying single molecules and estimating classification error in fluorescence microscopy images

Figure 4

The algorithm performs well even as data quality degrades. We tested the software on artificially constructed images of varying quality with realistic signal intensities, spot densities, and background noise based on real data. Spot density is measured by average distance of a pixel to an artificial spot in an image and is shown in the inset. Signal intensity on the x-axis is the average pixel intensity at the centers of the random spots minus the mean pixel intensity of the image divided by the standard deviation of the pixel intensities in the image. A sample of spots from the background 1 images were used to train the classifier.

Back to article page