| genre | Histogram |
| status | stable |
| author | Jan Rexilius |
| package | MeVisLab/Standard |
| dll | MLVolumetry |
| definition | MLVolumetry.def |
| keywords | histogram, peak, position, estimate |
The module HistogramPeakEstimate calculates an estimate of peak positions of a histogram.
The module computes peak positions of a smoothed version of the input histogram. Smoothing is performed by a non-linear diffusion filter with a linear diffusion model (see also SmoothHistogram).
A peak satisfies the following conditions:
- The histogram value of a peak is higher than those of his neighbors.
- The histogram value of a peak is at least Minimum Height (to remove low peaks).
- The difference between the histogram value of a peak and those of the two points with distance Distance for Slope is at least Minimum Slope (to remove flat peaks).
The positions of the peaks are returned, not the bin indices. All bin sizes are supported.
Example:
| Distance for Slope: Integer |
| Max Peak Postition: Integer |
| Min Peak Postition: Integer |
| Minimum Height: Float |
| Minimum Slope: Float |
| Number of Iterations: Integer |
| Number of Peaks: Integer |
| Time Step Size: Float |
Sets the number of iteration steps for the diffusion filter.
The more steps, the longer the diffusion time.
Sets the time step size.
Sets the minimum histogram value of a peak candidate to be considered.
Sets the minimum difference between the histogram value of a peak to be considered and those of the two points with distance Distance for Slope.
Sets the distance between histogram positions that is used for slope estimation.
Shows the number of computed peaks.