To display this dialog box, on the PRO tab, select Options and then click Image Processing.
Use the Image Processing dialog box to control how MapInfo Pro reprojects raster images.
Reprojection
Using
During the raster reprojection process, MapInfo Pro recalculates the pixel values of the source image to make them display correctly in the destination image. In this resampling process MapInfo Pro tries to restore every pixel value of the image based on the pixels around it. In MapInfo Pro there are two methods for calculating the pixel values of the destination image: Cubic Convolution and Nearest Neighbor. These are industry-standard terms used by GIS professionals all over the world.
Never
Select this options to indicate that Raster images should not be reprojected when added to a vector map and presents raster reprojection. This is the default setting which preserves the pre-version 8.5 MapInfo Pro behavior.
Always
Select this option to ensure that reprojection is always performed, that is, MapInfo Pro calculates the image's coordinates based on a precise formula and then the pixels are resampled using Cubic Convolution or Nearest Neighbor options.
Optimized
Select this option to ensure that raster image reprojection is determined by the look of the destination rectangle after transformation into the source image space. If it looks as a rigorous rectangle (two sides are parallel to x-axis and two sides parallel to y-axis), then the standard Windows functions stretch the source image in both directions as it was in pre-version 8.5 MapInfo Pro. If the image fails the rigorous rectangle test, the reprojection is performed using the specified resampling options.
Resampling Method
Cubic Convolution
This provides the best "restoration" of pixel values because of their discreetness. Using this option, a pixel in the destination image is calculated based on the pixel values in a 4x4 pixel window centered at the basic pixel in the source image. The coordinates of the basic pixel are calculated for every pixel of the destination image based on a special optimized procedure. Pixels are then weighted based on the basic pixel coordinates. In general, we recommend you use the Cubic Convolution resampling method for aerial images and satellite rasters to get a better image quality. The Cubic Convolution algorithm used in MapInfo Pro is based on the work of S.K. Park and R.A. Schowengerdt, Computervision, Graphics and Image Processing (1983, Volume 23. Pages 258-272).
Nearest Neighbor
This replaces the pixel value in the reprojected image with the basic pixel value from the source image. This resampling method takes less time to render than the Cubic Convolution method, but may be less precise. In general we recommend you use Nearest Neighbor resampling for raster maps, grids, and scanned maps to get faster results.