Stacking
Before proceding with the stacking process be sure all images have been aligned (see Images alignment) and then click the 'stack' button in the 'Tools' section.
A dialog window will be shown in which you can select the stcking method:
Currently the following stacking method are available:
- Average (native) : the average of all selected image is computed.
- Median : the median of all selected image is computed.
- Sigma-clipping : the average of all selected image is computed, and the standard deviation rejection method is applied [3].
- Standard deviation (native) : the standard deviation of all selected image is computed.
- Variance (native) : the variance of all selected image is computed..
- Maximum (native) : An image containing the maximum value for eache pixel (computed an all images) is returned.
- Minimum (native) : An image containing the minimum value for eache pixel (computed an all images) is returned.
- Product (native) : the product of all images is computed.
Here an example of the average of some images.
Once the stacking is completed, you can adjust the output level of the result image by clicking the 'Image levels' image in the 'Tools' section. The following dialog window will appear:
On the right side of the window the histograhm of the result image is displayed and on the left side there are the curves controls to adjust the levels of the image[4]. If the 'data clicking' section is enabled you can choose if to 'clip' pixel valuex outside the specified data range ( [0,255] or [0,65535] ) or to 'scale' the image levels in order that the maximum and minimum values are contained in the specified data range.
To explain how the curves work, we use the notation previously introduced in the section 'Zoom and Display'. So let I be the matrix of size WxHxC where W and H are the width and the height of the image and C the number of components. Now if I(x,y,c) is the element of I at position x,y,c and F(X) the curve to apply to I, then the result image I' is I'(x,y,c)=F(I(x,y,c)) for each x,y,c.
Now if we introduce the parameters A,B,O,M,N, the following curves can be defined:
- Linear: F(X) = A+B×X
- Logarithmic: F(X) = A + B×LogN(O + M×X)
- Power : F(X) = A + B×(O + M×X)N
- Exponential : F(X) = A + B×N(O + M×X)
The image shown below is the above one to which a Power curve is applyed.