Meziantou.FrameworkImageHistogram Class
An immutable histogram of one channel of the pixels of an image or a frame (ImageProcessingExtensions.ComputeHistogram): the number of pixels at each sample level.
public sealed class ImageHistogramRemarks
There is one bin per sample level, at the storage precision: 256 bins for 8-bit pixel formats, 65,536 for 16-bit ones. Every pixel is counted, fully transparent ones included.
Properties
| Name | Description |
|---|---|
| BinCount | Gets the number of bins: 256 for 8-bit pixel formats, 65,536 for 16-bit ones. |
| Channel | Gets the channel that is counted. |
| Counts | Gets the number of pixels of every level, indexed by the level. |
| TotalCount | Gets the number of pixels counted, the sum of the bins. |
BinCount
ImageHistogram.csChannel
ImageHistogram.csCounts
ImageHistogram.csGets the number of pixels of every level, indexed by the level.
public ReadOnlySpan<long> Counts { get; }Property value
TotalCount
ImageHistogram.csIndexers
| Name | Description |
|---|---|
| this[int] | Gets the number of pixels whose sample is a level. |
this[int]
ImageHistogram.csGets the number of pixels whose sample is a level.
Parameters
levelint- The level, from 0 to
BinCount - 1.
Property value
Exceptions
ArgumentOutOfRangeExceptionlevelis out of range.
Methods
| Name | Description |
|---|---|
| GetOtsuThreshold() | Gets the threshold that Otsu's method chooses: the level t that maximizes the variance between the class of the levels up to t and the class of the levels above it, w1 * w2 * (m1 - m2)², where w1 and w2 are the numbers of pixels of the classes and m1 and m2 their mean levels. |
GetOtsuThreshold()
ImageHistogram.csGets the threshold that Otsu's method chooses: the level t that maximizes the variance between the class of the levels up to t and the class of the levels above it, w1 * w2 * (m1 - m2)², where w1 and w2 are the numbers of pixels of the classes and m1 and m2 their mean levels.
public int GetOtsuThreshold()Returns
int The threshold: the pixels above it form the upper class.
Remarks
The candidates are the levels from the lowest one present to the one below the highest one present, and the first maximum wins. The variances are computed as scikit-image's threshold_otsu computes them, from exact integer sums, so the threshold is the one it returns for the same integer samples. When a single level is present, it is the threshold. ImageProcessingExtensions.Threshold with ThresholdMethod.Otsu uses the threshold of the HistogramChannel.Luminance histogram.