Computational image processing in microscopy

Abstract

Computational image processing provides a powerful tool to extract quantitative information from a large image data set. Images should be optimized for future processing during image acquisition. Pre-processing can remove noise and take other properties of the image into account in preparation for automated segmentation which identifies and delineates the objects in the image. Different segmentation methods are good for different types of images. Post-processing can reduce errors in segmentation using known properties of the objects identified such as size and distance between objects. It is important to validate the results and small errors can be hand corrected.