Here is a pre-release of the DefectMap process, which will be included in the next PI Core distribution, as part of the ImageCalibration module. As the ImageCalibration process, DefectMap is released as an open source project, subject to the GPL.
Brief decription: This process allows the user to use a reference image (called "Defect Map") to indicate bad pixels on a target image. The Defect Map may be a grayscale or color image, where 0 values means "bad pixel" (gray values are used as scaling factor in convolution modes, see below). So, for "graphical reference" it needs to look like a inverted dark frame. Also, it is assumed that the Defect Map needs no further processing, it is ready to use. If you are using a dark frame, or other calibration frame as defect map, you may use any other process to modify it.
Besides from the Defect Map image selection, the user may choose between five operation methods: Mean, Gaussian Blur, Median, Minimum and Maximum. Those operations needs a structuring element, defined by the user following any of these geometries: Square, Circular, Row or Column. The size of the operation is automatically calculated for every bad pixel. Of course, no bad pixels are included in any of this calculations, regardless of the structuring element selected. Pixel values above 0 on the Defect Map act as a weight factor for the Mean and Gaussian blur operators.
Bayered images are not directly supported. Anyway, It is possible to modify/build the Defect Map in such a way that it uses pixels that corresponds to a certain color only.
Download links:
Source code:
http://pteam.pixinsight.com/pcldev/defectmap/DefectMap-src-20100504.zipWindows 32bits:
http://pteam.pixinsight.com/pcldev/defectmap/DefectMap-pxm-20100504_win_32.zipWindows 64bits:
http://pteam.pixinsight.com/pcldev/defectmap/DefectMap-pxm-20100504_win_64.zipLinux 64bits:
http://pteam.pixinsight.com/pcldev/defectmap/DefectMap-pxm-20100504_x11_64.tar.gz(Will include 32bits releases if asked).
Feedback is appreciated
