| mosaic |
Mosaic parameter-file-name
The mosaic program supports 4 different kinds of mosaics:-
- Simple Mosaics(Mosaic Type 0) These are the default mosaics created simply by transforming the source images into the target image space. There is no contrast or seam correction. (This takes the minimum execution time)
- Mosaics with Averaging (Mosaic Type 1)These type of mosaics try to reduce sharp edge transitions by averaging the pixel values which fall within the same position in the target image space.(These also take minimal execution overhead)
- Mosaics with Feathering (Mosaic Type 2)These type of mosaics try to reduce sharp edge transitions by using a weighted average of source images where the weights are determined by a hat function.A blending factor can be specified to control the degree of blending(refer optional parameters) .These type of mosaics run slower compared to the Mosaic types 0 and 1.
- Mosaics with Feathering and Contrast Correction (Mosaic Type 3)These type of mosaics additionally perform contrast correction by computing a relative contrast gain between each source image relative to the target.This is a 2 pass algorithm and takes twice the time taken to perform feathering alone.The contrast correction allows the user to specify a "target contrast" The contrast of the mosaiced image will be adjusted to match the target contrast image.
Additional (Optional Parameters)
- MOSAIC_TYPE:
This is an integer value which specifies the type of mosaic to be generated . Possible values (0,1,2,3)
- BLEND_FACTOR:
This is a floating point value which is used in calculating weights for the feathering calculations.(Low values will give results similar to averaging .Increasing the value causes each source image's contribution to be restricted to pixels closest to its center.(A value of 2-3 will give good results)
- TARGET_CONTRAST:
This is an integer value which specifies the image to which the contrast will be adjusted.For example:If target contrast is unspecified the contrast/brightness of the mosaiced image would be adjusted to match that of the target image.
|