overlap kernels strategies

There is several strategies possible to define how to stich overlaping areas. On the following points you will see a short descrition of each strategy and an example of how it will stitch images together and ‘expected’ results. Here are the ‘raw’ images used for this example:

stitching raw data

mean strategy

It will sum the mean value of the two images

mean strategy

closest strategy

It will take the top image for the top part and the bottom image for the bottom part

closest strategy

linear weights strategy

It will decrease lineary the weights of the top frame and increase the effect of the bottom frame.

linear strategy

cosinus weights strategy

This one is close to the linear weights strategy except that instead of having a linear profile it will have a ‘cosinus / sinus’ profile and should insure a smoother transition.

linear strategy