Binarization of Inhomogeneously Illuminated Images

Most grey level thresholding methods produce good results only when the illumination of the image is homogeneous. An automatic binarization technique suitable for images containing regions of different brightness is presented. It does not use image’s grey level histogram as a source of information. Rather, the method is based on the transformation of the raster image in such a way that the transformed image can be easy thresholded. A comparison with histogram based threshold selection technique is given.

• Type: Conference
• Year: 1995
• Where: ICIAP
• Authors: Volodymyr V. Kindratenko, Boris A. Treiger, Piet J. M. Van Espen
• Keywords: Grey Level, Grey Level Histogram, ICIAP 1995, Image Processing, Most Grey Level