find_peaks#

gammapy.estimators.utils.find_peaks(image, threshold, min_distance=1)[source]#

Find local peaks in an image.

This is a very simple peak finder, that finds local peaks (i.e. maxima) in images above a given threshold within a given min_distance around each given pixel.

If you get multiple spurious detections near a peak, usually it’s best to smooth the image a bit, or to compute it using a different method in the first place to result in a smooth image. You can also increase the min_distance parameter.

The output table contains one row per peak and the following columns:

  • x and y are the pixel coordinates (first pixel at zero)

  • ra and dec are the RA / DEC sky coordinates (ICRS frame)

  • value is the pixel value

It is sorted by peak value, starting with the highest value.

If there are no pixel values above the threshold, an empty table is returned.

There are more featureful peak finding and source detection methods e.g. in the photutils or scikit-image Python packages.

Parameters
imageWcsNDMap

Image like Map

thresholdfloat or array-like

The data value or pixel-wise data values to be used for the detection threshold. A 2D threshold must have the same shape as the map data.

min_distanceint or Quantity

Minimum distance between peaks. An integer value is interpreted as pixels.

Returns
outputTable

Table with parameters of detected peaks