IPP Software Navigation Tools IPP Links Communication Pan-STARRS Links

Ignore:
Timestamp:
Aug 28, 2008, 6:04:19 PM (18 years ago)
Author:
Paul Price
Message:

We want a bad/poor mask distinction in the mask convolution for stacking. Again, otherwise the number of bad pixels just explodes. Since convolving a mask with the PSF-matching kernel is something we want to do in both stack rejection and normal convolution, pulled the code that determines the size of the 'bad' region into a separate function in pmSubtraction.c. Then pmStackReject basically does the same thing that's happening in pmSubtractionConvolve/convolveRegion: go through the image by footprints, convolve the footprint and plug the result back in to the original image. Small API change for pmStackReject. This is very effective in reducing the amount of bad pixels in the stacked image.

File:
1 edited

Legend:

Unmodified
Added
Removed
  • trunk/psModules/src/imcombine/pmStackReject.h

    r16607 r19282  
    1010/// We apply a matched filter to the corresponding mask image, and threshold to find the original pixels
    1111psPixels *pmStackReject(const psPixels *in, ///< List of pixels in the convolved image
    12                         const psRegion *valid, ///< Valid region to consider
     12                        int numCols, int numRows, ///< Size of image of interest
    1313                        float threshold, ///< Threshold on convolved image, 0..1
     14                        float poorFrac, ///< Fraction for "poor"
    1415                        const psArray *regions, ///< Array of image regions for image
    1516                        const psArray *kernels ///< Array of kernel parameters for each region
Note: See TracChangeset for help on using the changeset viewer.