Automatic extracted object technique for contrast enhancement medical images

Abstract EN

The aim of this work is a proposed system to enhance automatically the contrast of the desired region in the medical image to get the wanted information without enhancing the contrast of the whole image.

The proposed system includes the automatic extraction process, the automatic contrast enhancement process and finally reinsertion process for region in the image.

Threshold, smoothing, boundary extraction, chain code or region filling techniques were used in the automatic extraction process.

While, histogram equalization, histogram stretching, gray level grouping (GLG), fast gray level grouping (FGLG) or auto-contrast techniques were used in the automatic contrast enhancement process for an extraction region from the image.

According to image appearance, GLG, FGLG and histogram stretching are the best techniques to enhance contrast in the RGB image.

Auto-contrast technique lowers accuracy, while histogram equalization results are unacceptable.