PRACTICAL METHODS FOR DIGITAL IMAGE-ENHANCEMENT IN SEM

In microscopy, the most important image information is found in image details, i.e., the smallest image features. As an initial step in imag e detail evaluation, the whole image must be surveyed for relevant inf ormation. Then, identified features may be further inspected. Two task s must be accomplished in a first approach to full-frame image evaluat ion: the overall contrast must be balanced for assessment of large ima ge features, and the local contrasts must be enhanced for assessment o f detail features. A new automatic ''highlight'' filter proved very us eful for the achievement of both tasks. The filter makes it possible t o visually access the image detail content immediately after image acq uisition and thus facilitates microscopy on difficult samples. The dig ital filter operates automatically without the need for specific param eters under normal operating conditions. A median tilter with an unusu ally large mask size is used for the extraction of detail contrasts ch aracterized as highlight contrast, and histogram equalization is used for the enhancement of the highlight contrasts. The processing utilize s digital signal processors for high-speed image enhancement facilitat ing visual analysis of the improved image. The filter's performance is compared with that of conventional processing techniques and microsco pic image acquisition under varied conditions. The highlight filter fa cilitates immediate assessment of image detail content and contrast me chanisms, and thus improves routine microscopy.