The H-Minima Transform was invented to reduce the over-segmentation of the watershed algorithm by reducing noise in the darker regions of the gray value image.

The method searches for local minima in the gray value image and suppresses them, if their depth is lower or equal than the Minimum Height. The depth of a local minimum is defined as the difference between the gray value in the local minimum and the next local gray value maximum.

In the simplified 2D example below, the neighboring gray values are shown in gray bars. For each local minimum the corresponding depth is visualized in blue.

For comparison, the given Minimum Height is visualized as a red rectangle. Observe that the depth of the two minima in the center is lower than the Minimum Height, while the other two have a depth higher than the Minimum Height. Thus, the two in the middle will get suppressed and the other two will not, as shown below.

The Tolerance is the relative change in the image mean gray value. If 0 is entered for Tolerance the runtime can be very long, trying to find a perfectly stable solution. In most cases, however, a slightly higher tolerance, for example the default of 0.001, already leads to good results.

Click Apply to apply the filter on the gray value image. In the following, the example was already filtered with the Non-Local Means Filter and a Morphological Gradient. Then, the H-Minima Transform is applied with a Minimum Height of 30.