In this section, we present further images which are enhanced by means of our min/max flows. We begin with a series of medical images in Figure 13; here, no noise is artificially added, and instead our goal is to enhance certain features within the given images.
Figure 13: Min/Max Flow with Selective Smoothing
Next, we study the effect of our min/max scheme on multiplicative noise
added to a grey-scale image. In Figure 14 we show the
reconstruction of an image with 
multiplicative noise.
Figure 14: Min/Max Flow Applied to Multiplicative Noise
Next, we add 
Gaussian grey-scale noise; that is, a random
component drawn from a Gaussian distribution with mean zero is
added to each (every) pixel. In Figure 15 we show the
original with noise together with the reconstructed min/max flow image. Here
we use the coupling 
.
Figure 15: Continuous Gaussian Noise added to Image
Finally, we apply our scheme to a color image with
Gaussian color noise added; that is, Gaussian noise added to
all three channels.
In Figure 16(c) we show the original with noise and in Figure
16(d) we show the image reconstructed using min/max flow.
Figure 16: Reconstruction of color images using Min/Max flow:
(This is a 24-bit
color image; quality is degraded if grey-scale reproduced.)