Image Health

Image Health (IH) is a python module used by the image builders to perform quick data quality tests on each image in real-time.


For data quality assurance purposes, IH has the current functionality:

  • Calculates the mean (bias level), standard deviation (bias noise), and slope of the pixels in the horizontal and vertical overscan regions of each amplifier of each CCD
  • Calculates the median (bias level + sky background) and data noise values for the pixels in the data regions of each amplifier of each CCD. IH uses a sigma-clipping algorithm to reject outlying pixels in the noise calculation
  • Calculates astrometric and photometric quantities for acceptable stars in each amp of each CCD (excluding the focus and alignment CCDs). IH does not use any pattern recognition algorithms to separate between galaxies and stars; rather, a set of simple tests are performed that confirm if the final objects used for calculations are stars. These tests include:
    • The object contains no saturated pixels
    • The perimeter of the search box (centered on the brightest object pixel) around the object contains no outlying pixels (i.e., the object is isolated)
    • The object is above a minimum area and below a maximum area in size
  • Calculates the following quantities for each star (units in brackets):
    • x barycenter [pixel coordinate]
    • y barycenter [pixel coordinate]
    • x brightest [pixel coordinate]
    • y brightest [pixel coordinate]
    • variance along x [pixel coordinate^2]
    • variance along y [pixel coordinate^2]
    • covariance of x and y [pixel coordinate^2]
    • rotation angle [deg]
    • semi-major axis [pixel]
    • semi-minor axis [pixel]
    • ellipticity
    • isophotal flux (background subtracted) [ADU]
    • full width at half-max [arcsec]
  • Presents calculated quantities in an organized, easy-to-read format using custom graphical user interfaces (GUIs)
  • Loads alarm thresholds from the constants database (DB). Generates warnings/alarms and stores them in the alarm DB
  • Posts results as telemetry in the DB. Various plots, including history plots, of these results may be viewed using the online telemetry viewer

Image Health Details [SISPI Wiki]

Image Health Presentation [6/14/2012, docDB #6544]