Abstract: A histogram modeling based technique for image contrast enhancement. In some implementations, a histogram of an image is created and then transformed. Using the physics of sound or heat propagation, the technique may develop a spreaded histogram model that may be transformed. A nonlinear mapping may be created to remap an image for contrast enhancement. The technique may be performed without threshold tuning and may be implemented on a variety of display hardware.

Abstract: There is provided an image processing apparatus such as facsimile apparatus or digital copying machine for digitizing image data into two values or multi-values. The apparatus comprises: an input device to input data of a target pixel to be digitized; a calculating circuit to obtain an average density value of a predetermined region; a digitizing circuit such as a binarizing circuit for digitizing or binarizing the target pixel data on the basis of the average density value obtained by the calculating circuit; and a correction circuit to correct a difference which occurs upon digitization by the digitizing circuit. The calculating circuit calculates the average density value m.sub.1 of the pre-determined region when the target pixel data was binarized to 1 and the average density value m.sub.2 of the predetermined region when the target pixel data was binarized to 0 by using a predetermined weighting mask. The digitizing circuit binarizes the target pixel data by using (m.sub.1 +m.sub.