Sometimes there are homogeneous areas in a raster dataset that you do not want to display. These can include borders, backgrounds, or other data considered to not have valid values. Sometimes these are expressed as NoData values, although at other times they may have real values.
Backgrounds and outlines can often be the result of georeferencing your raster dataset. If your raster data has a background, border, or other NoData values, you can choose not to display them or choose to display them as a particular color.
All renderers allow you to set the NoData value to a color or No Color, while the Stretched renderer allows you to identify a specific background value and display color or No Color.
What do you do if you’re still seeing a background colour after applying a 255,255,255 RGB background and the NoData options as in the example below?
If you had other values other than 255, 255, 255 for colours close to white (like 250 for example) and the NoData and Background colour settings were not enough to handle them, then it’s likely that that they may otherwise be rendering errors coming from the image preprocessing, perhaps from compressing the images.
Trying to apply the Spatial Analyst > Reclass > Reclassify tool try to handle that problematic range would only result in you losing the detail in your image by simplifying the spectrum into a new classification.
A better solution is a workflow that should make your data management more efficient as a consequence.
You will need to create a Mosaic Dataset to manage your images and be able to specify the reclassification this way.
The mosaic dataset simply acts as a reference to your images to spatially index them with pyramids for processing and any queries or conditions (like the reclassification of RGB values > 250 for example) you may want to add to them during display.