Combine Images
The Combine Images command combines the images in an Image Set into a single new image. Applications include stacking images after registration, creating master calibration frames, and creating high signal-to-noise ratio images. This command is not used to merge color channels into an RGB image.
Open this dialog using the from the Process > Math Menu.
Combine Images uses an Image Set displayed in an Image Window. The result is an output image created in a new window. Several statistical methods are available for combining the images and for adjusting ("normalizing") them to a common reference value (see Image Combining Methods and Region Statistics Methods). The settings are saved as a Profile.
These settings appear when the dialog is in the "dropped", or open position.
Method |
This list box selects the type of processing used to combine the image values. These are described in Image Combining Methods. Depending upon the method selected, extra fields may appear for method-specific parameters, like High sigma. |
Normalization |
This control group specifies whether the images are adjusted (normalized) to a common statistical value before being combined. If normalization is not set to "none", then the button opens the Region Stat Preferences dialog. In this dialog you specify how the normalization is to be computed, including the target statistic and the region over which it is calculated. See the discussion below. |
Data Type |
Select the Data Type so that it can accommodate the increased signal or dynamic range of the combined images. For example, if combining a large number of 16-bit integer type images it is usually prudent to set the Data type of the output image to 32 bit real. |
Other Options |
These are miscellaneous settings grouped together to conserve screen space:
|
Normalization is used to adjust the images to a common statistical value when using any combining method that rejects bad data. This is essential when using a combining method that performs rejection of bad data, such as the Sigma Clipping, Min/Max or Median methods.
Choose the normalization operation, Scale, Offset, or None, according to the nature of the difference between the images being combined.
If the pixel values differ because of an additive offset, like variations in the CCD bias offset, select Offset.
If the images differ because of a scale factor, as caused by differences in illumination or exposure time, select Scale.
The None setting disables normalization.
Math Commands, Image Set, Image Combining Methods, Region Statistics Methods, Region Stat Preferences, Register Images, Drop Dialogs, Profiles