Balancing areas of varying density in a digital image

Abstract

A method of producing a mask for digital unsharp masking or for digital equalization and a computer readable memory having recorded thereon statements and instructions for execution on a computer for carrying out the method both comprise: receiving an unprocessed digital radiograph; creating a first new image that is a copy of the radiograph; then creating a second new image by blurring the first new image with a kernel of a selected starting kernel size.

1 . A method of producing a mask for digital unsharp masking or for digital equalization, comprising: 
 (a) selecting a starting kernel size;    (b) selecting a minimum kernel size;    (c) selecting a number of iterations;    (d) receiving an unprocessed digital radiograph;    (e) creating a first new image by copying the radiograph;    (f) creating a second new image by blurring the first new image with a kernel of the starting kernel size;    (g) replacing pixel values in the first new image by retrieving the pixel-wise maximum of the second new image and the radiograph;    (h) dividing the starting kernel size by two; then    (i) repeating steps (f) through (h) for either the selected number of iterations or until the starting kernel size is less than or equal to the minimum kernel size.    
     
     
         2 . A method as claimed in  claim 1  wherein in step (a) the starting kernel size is between 100×100 and 200×200 pixels.  
     
     
         3 . A method as claimed in  claim 1  wherein in step (f), the second new image has a blurred intensity determined for each pixel in the second new image by referencing an unequalized intensity value of the pixels in a kernel surrounding each pixel.  
     
     
         4 . A method of  claim 3  wherein the blurred intensity for each pixel in the second new image is the average of the unequalized intensity values of the pixels in the kernel surrounding each pixel.  
     
     
         5 . A method of producing a mask for digital unsharp masking or for digital equalization, comprising: 
 (a) receiving an unprocessed digital radiograph;    (b) creating a first new image that is a copy of the radiograph; then    (c) creating a second new image by blurring the first new image with a kernel of a selected starting kernel size.    
     
     
         6 . A method as claimed in  claim 5  further comprising after (c): 
 (d) replacing pixel values in the first new image by retrieving the pixel-wise maximum of the second new image and the radiograph;    (e) reducing the starting kernel size; then    (f) repeating steps (c) to (e) until the starting kernel size is less than or equal to a selected number of iterations or a selected minimum kernel size.    
     
     
         7 . A method as claimed in  claim 6  wherein in (e) the starting kernel size is reduced by two.  
     
     
         8 . A computer-readable memory having recorded thereon statements and instructions for execution by a computer to carry out a method for producing a mask for digital unsharp masking or for digital equalization comprising: 
 (a) selecting a starting kernel size;    (b) selecting a minimum kernel size;    (c) selecting a number of iterations;    (d) receiving an unprocessed digital radiograph;    (e) creating a first new image by copying the radiograph;    (f) creating a second new image by blurring the first new image with a kernel of the starting kernel size;    (g) replacing pixel values in the first new image by retrieving the pixel-wise maximum of the second new image and the radiograph;    (h) dividing the starting kernel size by two; then    (i) repeating steps (f) through (h) for either the selected number of iterations or until the starting kernel size is less than or equal to the minimum kernel size.    
     
     
         9 . A memory as claimed in  claim 8  wherein in step (a) the starting kernel size is between 100×100 and 200×200 pixels.  
     
     
         10 . A memory as claimed in  claim 9  wherein in step (f), the second new image has a blurred intensity determined for each pixel in the second new image by referencing an unequalized intensity value of the pixels in a kernel surrounding each pixel.  
     
     
         11 . A memory as claimed in  claim 7  wherein the blurred intensity for each pixel in the second new image is the average of the unequalized intensity values of the pixels in the kernel surrounding each pixel.  
     
     
         12 . A computer-readable memory having recorded thereon statements and instructions for execution by a computer to carry out a method for producing a mask for digital unsharp masking or for digital equalization comprising: 
 (a) receiving an unprocessed digital radiograph;    (b) creating a first new image that is a copy of the radiograph; then    (c) creating a second new image by blurring the first new image with a kernel of a selected starting kernel size.    
     
     
         13 . A memory as claimed in  claim 12  further comprising after (c): 
 (d) replacing pixel values in the first new image by retrieving the pixel-wise maximum of the second new image and the radiograph;    (e) reducing the starting kernel size; then    (f) repeating steps (c) to (e) until the starting kernel size is less than or equal to a selected number of iterations or a selected minimum kernel size.    
     
     
         14 . A memory as claimed in  claim 13  wherein in (e) the starting kernel size is reduced by two.