Most LEADTOOLS method for image manipulation and analysis act on the image in memory. It is important to keep this in mind, because the changes become permanent when you save the image in a file. Sometimes you want the changes to be permanent, and sometimes you do not. It all depends on the reason for making the changes.
Sometimes, you start with an image that has some undesirable qualities. For example, if the original image is too flat, you may want to use one of the contrast enhancement methods to make permanent improvements.
At other times, you may need to make temporary changes for use on a particular monitor or printer. Using the .NET methods, you can set the brightness, contrast, and gamma correction for the display of all images, without affecting the image data. But for other changes, such as creating a halftone for printing, a good strategy is to modify a copy of the image, and give the user the ability to undo changes. The undo strategy also makes sense with a manipulation, such as a mosaic effect, that intentionally distorts an image.
Some image manipulations, such as edge-detection filters, are commonly used to analyze an image, rather than to improve it. For example, in an industrial quality-assurance application, defects in a machine part may be easier to detect if an image of the part is processed with an edge-detection filter. Similarly, the creation of a histogram (an array that charts the frequency of color usage in an image) can be of value in an analytical situation.
Edge detection and histogram creation can also be used as steps in improving an image. For example, creating a histogram is a necessary step, internally, in some of the contrast-improvement methods. Also, an image created with an edge detection filter can be combined with the original image to harden or soften the original's edges.