Method and apparatus for effecting background suppression of image data

Claim 1

Apparatus for processing an image containing image formation data and background, the image being represented by an array of pixels, each pixel initially having a gray-scale value in a first preselected range of values, said apparatus comprising: background suppression means for converting each pixel value to preselected values within a second range of values smaller than said first preselected range, said background suppression means including: a) background reference threshold means for providing a background reference threshold value; b) dynamic threshold means for providing a dynamic threshold value which is different from the background threshold value; the dynamic threshold means comprising gradient means for generating at least one gradient value as a function of the difference between values of neighboring pixels, wherein the gradient means operates on pixels in a window and generates a current horizontal gradient, a current vertical gradient, a current gradient magnitude, a previous horizontal gradient, a previous vertical gradient, a previous gradient magnitude, and an average local gray value, the dynamic threshold means further comprising histogram counter means for generating a histogram of said gradient values, gradient threshold select means for selecting a gradient threshold value which is a function of said histogram values, and look up table means for generating said dynamic threshold value as a function of the gradient threshold value; c) enabler/selector means for analyzing predetermined patterns of pixel values surrounding each pixel being converted and selecting the dynamic threshold value or background reference threshold value as a function of the analysis; the enabler/selector means comprising stroke edge detector means for detecting a change in local contrast in the image before and after each pixel, said stroke edge detector means providing an output which is used to determine whether the dynamic threshold value or background reference threshold value is selected; and d) remapper means for assigning a new value in said second range to each pixel in the image as a function of the relationship between the initial gray-scale pixel value and the selected threshold value.

Claim 11

A method for processing an image containing image formation data and background, the image being represented by an array of pixels, each pixel initially having a gray-scale value in a first preselected range of values, said method comprising: (a) selecting a background reference threshold value; (b) selecting a dynamic threshold value which is different from the background threshold value by (i) generating at least one gradient value as a function of a difference between values of neighboring pixels, (ii) generating a histogram of said gradient values, and (iii) selecting a gradient threshold value by detecting a peak in the histogram and generating a gradient threshold value as a function of the peak and a user programmable factor; (c) analyzing a predetermined pattern of pixel values surrounding each pixel being processed; (d) selecting the dynamic threshold values or said background reference value as a function of the analysis; and (e) assigning a new gray-scale value for each pixel, said new gray-scale value lying within a second range of values which is smaller than said first preselected range of values, to each pixel in the image as a function of the relationship between the initial gray-scale pixel value and the value selected in step (d).

Claim 12

A method for processing an image containing image formation data and background, the image being represented by an array of pixels, each pixel initially having a gray-scale value in a first preselected range of values, said method comprising: (a) selecting a background reference threshold value; (b) selecting a dynamic threshold value which is different from the background threshold value by (i) generating at least one gradient value as a function of a difference between values of neighboring pixels, (ii) generating a histogram of said gradient values, and (iii) selecting a gradient threshold value by detecting a peak in the histogram and generating a gradient threshold value as a function of the peak and a user programmable factor; (c) analyzing a predetermined pattern of pixel values surrounding each pixel being processed; (d) selecting the dynamic threshold values or said background reference value as a function of the analysis; and (e) assigning a new gray-scale value for each pixel, said new gray-scale value lying within a second range of values which is smaller than said first preselected range of values, to each pixel in the image as a function of the relationship between the initial gray-scale pixel value and the value selected in step (d).