Automatic tone correction of images using non-linear histogram processing

• US 5,544,258 A
• Filed: 07/12/1993
• Issued: 08/06/1996
• Est. Priority Date: 03/14/1991
• Status: Expired due to Fees

• Current Assignee: Raphael L. Levien
• Original Assignee: Raphael L. Levien
• Inventors: Levien, Raphael L.
• Primary Examiner(s): Boudreau, Leo
• Assistant Examiner(s): Johns, Andrew W.
• Application Number:
  US08/089,622
• Time in Patent Office: 1,121 Days
• Field of Search: 382/51, 382/54, 382/169, 382/274, 358/163, 358/168, 358/169, 358/448, 358/455, 358/461, 358/466, 358/522, 358/523, 358/524
• US Class Current: 382/169
• CPC Class Codes: G06T 5/20 using local operators
• CPC Class Codes: G06T 5/40 using histogram techniques
• CPC Class Codes: G06T 5/92 based on global image properties

Abstract

A method and apparatus for automatically correcting the tone values of a digitally stored image includes compiling a histogram of the relative occurrence of gray values in the stored image in memory. The histogram is processed by adding a constant, taking the square root, and digital filtering. The processed histogram is numerically integrated and then normalized to form a tone correction function. The present tone correction method may be used to tone correct color images by applying separate tone correction functions generated for each of the three color planes independently, or in the alternative, by applying a composite tone correction function corresponding to the average of histograms of the respective individual color planes. A switch is provided to allow the operator to view the results of both methods of automatic tone correction for color images, and choose the best resulting image.

Claims

1. In an automatic tone correction system wherein a stored image plane including a plurality of pixels, is processed by applying a tone correction function to each pixel of said stored image plane, a method of generating a tone corrected image from said stored image, comprising the steps of:

compiling a histogram of the relative occurrence of tone values corresponding to said stored pixels of said stored image plane;

processing said histogram with a non-linear function so that small histogram values have relatively more contribution to said tone correction function, and large histogram values have relatively less contribution to said tone correction function, said non-linear function comprising raising each histogram value to a power between 0 and 1;

numerically integrating said processed histogram;

normalizing said numerically integrated histogram to extend over the desired range of tone values in interest in order to form said tone correction function;

storing said tone correction function in a lookup table memory; and

applying said stored plurality of pixels of said stored image plane in said memory to said lookup table memory to provide a plurality of tone corrected image output pixels from said lookup table memory corresponding to said tone corrected image.

2. A method in accordance with claim 1, further comprising:

filtering said histogram to smooth out the peaks of said histogram.

3. A method in accordance with claim 1, further comprising:

adding a constant to said histogram.

4. In an automatic tone correction system wherein a stored image plane including a plurality of pixels, is processed by applying a tone correction function to each pixel of said stored image plane, a method of generating a tone corrected image from said stored image, comprising the steps of:

compiling a histogram of the relative occurrence of tone values corresponding to said stored pixels of said stored image plane;

processing said histogram with a non-linear function so that small histogram values have relatively more contribution to said tone correction function, and large histogram values have relatively less contribution to said tone correction function, wherein said processing step further comprises computing the square root of each of said histogram values;

numerically integrating said processed histogram;

normalizing said numerically integrated histogram to extend over the desired range of tone values in interest in order to form said tone correction function;

storing said tone correction function in a lookup table memory; and

applying said stored plurality of pixels of said stored image plane in said memory to said lookup table memory to provide a plurality of tone corrected image output pixels from said lookup table memory corresponding to said tone corrected image.

5. In an automatic tone correction system wherein a stored image plane is processed by applying a tone correction function to each pixel of said stored image plane, a method of generating a tone corrected image from said stored image, comprising the steps of:

scanning a photographic image to provide a plurality of pixels corresponding to an image plane;

storing said plurality of pixels of said image plane in a memory;

compiling a histogram of the relative occurrence of tone values corresponding to said stored pixels of said stored image plane;

processing said histogram with a non-linear function so that small histogram values have relatively more contribution to said tone correction function, and large histogram values have relatively less contribution to said tone correction function;

numerically integrating said processed histogram;

normalizing said numerically integrated histogram to extend over the desired range of tone values in interest in order to form said tone correction function;

filtering said histogram to smooth out the peaks of said histogram, wherein said step of filtering said histogram further comprises;

filtering said histogram in a first low pass filter;

storing the output of said first low pass filter in a first time reversal buffer;

filtering the output of said first time reversal buffer in a second low pass filter; and

storing the output of said second low pass filter in a second time reversal buffer;

storing said tone correction function in a lookup table memory;

applying said stored plurality of pixels of said image plane in said memory to said lookup table memory to provide a plurality of tone corrected image output pixels from said lookup table memory corresponding to a tone corrected image; and

displaying said tone corrected image on a display device.

6. In an automatic tone correction system wherein a stored image plane including a plurality of pixels, is processed by applying a tone correction function to each pixel of said stored image plane, a method of generating a tone corrected image from said stored image, comprising the steps of:

compiling a histogram of the relative occurrence of tone values corresponding to said stored pixels of said stored image plane;

processing said histogram, said histogram processing comprising;

adding a constant to said histogram;

processing said histogram with a non-linear function so that small histogram values have relatively more contribution to said tone correction function, and large histogram values have relatively less contribution to said tone correction function, said non-linear function comprising raising each histogram value to a power between 0 and 1; and

filtering said histogram to smooth out the peaks of said histogram;

numerically integrating said processed histogram;

normalizing said numerically integrated histogram to extend over the desired range of tone values in interest in order to form said tone correction function;

storing said tone correction function in a lookup table memory; and

applying said stored plurality of pixels of said stored image plane in said memory to said lookup table memory to provide a plurality of tone corrected image output pixels from said lookup table memory corresponding to said tone corrected image.

7. In an automatic tone correction system wherein a stored image plane including a plurality of pixels, is processed by applying a tone correction function to each pixel of said stored image plane, a method of generating a tone corrected image from said stored image, comprising the steps of:

compiling a histogram of the relative occurrence of tone values corresponding to said stored pixels of said stored image plane;

processing said histogram, said histogram processing comprising;

adding a constant to said histogram;

processing said histogram with a non-linear function so that small histogram values have relatively more contribution to said tone correction function, and large histogram values have relatively less contribution to said tone correction function comprising computing the square root of each tone value; and

filtering said histogram to smooth out the peaks of said histogram;

numerically integrating said processed histogram;

normalizing said numerically integrated histogram to extend over the desired range of tone values in interest in order to form said tone correction function;

storing said tone correction function in a lookup table memory; and

applying said stored plurality of pixels of said stored image plane in said memory to said lookup table memory to provide a plurality of tone corrected image output pixels from said lookup table memory corresponding to said tone corrected image.

8. In an automatic tone correction system wherein a color image is stored in the form of a plurality of color planes, said color image being processed by applying a tone correction function to each pixel of said stored color planes, a method of generating a tone corrected image from said stored image, comprising the steps of:

compiling a histogram of the relative occurrence of tone values corresponding to said stored pixels of each of said color planes of said color image;

averaging said plurality of histograms to form a composite histogram;

processing said composite histogram with a non-linear function so that small histogram values have relatively more contribution to said tone correction function, and large histogram values have relatively less contribution to said tone correction function, said non-linear function comprising raising each histogram value to a power between 0 and 1;

numerically integrating said processed histogram;

normalizing said numerically integrated histogram to extend over the desired range of tone values in interest in order to form said tone correction function;

storing said tone correction function in a lookup table memory; and

applying said stored plurality of pixels of one of said color planes in said memory to said lookup table memory to provide a plurality of tone corrected image output pixels from said lookup table memory corresponding to said tone corrected image.

9. In an automatic tone correction system wherein a color image is stored in the form of a plurality of color planes, said color image being processed by applying a tone correction function to each pixel of said stored color planes, a method of generating a tone corrected image from said stored image, comprising the steps of:

compiling a histogram of the relative occurrence of tone values corresponding to said stored pixels of each of said color planes of said color image;

averaging said plurality of histograms to form a composite histogram;

processing said composite histogram with a non-linear function so that small histogram values have relatively more contribution to said tone correction function, and large histogram values have relatively less contribution to said tone correction function, wherein said processing step further comprises computing the square root of each tone value;

numerically integrating said processed histogram;

normalizing said numerically integrated histogram to extend over the desired range of tone values in interest in order to form said tone correction function;

storing said tone correction function in a lookup table memory; and

applying said stored plurality of pixels of said color plane in said memory to said lookup table memory to provide a plurality of tone corrected image output pixels from said lookup table memory corresponding to said tone corrected image.

10. In an automatic tone correction system wherein a stored image plane including a plurality of pixels, is processed by applying a tone correction function to each pixel of said stored image plane, an apparatus for generating a tone corrected image from said stored image, comprising:

means for compiling a histogram of the relative occurrence of tone values corresponding to said stored pixels of said stored image plane;

means for processing said histogram with a non-linear function so that small histogram values have relatively more contribution to said tone correction function, and large histogram values have relatively less contribution to said tone correction function, said non-linear function comprising raising each histogram value to a power between 0 and 1;

means for numerically integrating said processed histogram;

means for normalizing said numerically integrated histogram to extend over the desired range of tone values in interest in order to form said tone correction function;

means for storing said tone correction function in a lookup table memory; and

means for applying said stored plurality of pixels of said stored image plane in said memory to said lookup table memory to provide a plurality of tone corrected image output pixels from said lookup table memory corresponding to said tone corrected image.

11. An apparatus in accordance with claim 10, further comprising:

means for filtering said histogram to smooth out the peaks of said histogram.

12. An apparatus in accordance with claim 10, further comprising:

means for adding a constant to said histogram.

13. In an automatic tone correction system wherein a stored image plane including a plurality of pixels, is processed by applying a tone correction function to each pixel of said stored image plane, an apparatus for generating a tone corrected image from said stored image, comprising:

means for compiling a histogram of the relative occurrence of tone values corresponding to said stored pixels of said stored image plane;

means for processing said histogram with a non-linear function so that small histogram values have relatively more contribution to said tone correction function, and large histogram values have relatively less contribution to said tone correction function, said means for processing said histogram further comprises means for computing the square root of each of said histogram values;

means for numerically integrating said processed histogram;

means for normalizing said numerically integrated histogram to extend over the desired range of tone values in interest in order to form said tone correction function;

means for storing said tone correction function in a lookup table memory; and

means for applying said stored plurality of pixels of said stored image plane in said memory to said lookup table memory to provide a plurality of tone corrected image output pixels from said lookup table memory corresponding to said tone corrected image.

14. In an automatic tone correction system wherein a stored image plane is processed by applying a tone correction function to each pixel of said stored image plane, an apparatus for generating a tone corrected image from said stored image, comprising:

means for scanning a photographic image to provide a plurality of pixels corresponding to an image plane;

means for storing said plurality of pixels of said image plane in a memory;

means for compiling a histogram of the relative occurrence of tone values corresponding to said stored pixels of said stored image plane;

means for processing said histogram with a non-linear function so that small histogram values have relatively more contribution to said tone correction function, and large histogram values have relatively less contribution to said tone correction function;

means for numerically integrating said processed histogram;

means for normalizing said numerically integrated histogram to extend over the desired range of tone values in interest in order to form said tone correction function;

means for filtering said histogram to smooth out the peaks of said histogram, wherein said means for filtering said histogram further comprises;

means for filtering said histogram in a first low pass filter;

means for storing the output of said first low pass filter in a first time reversal buffer;

means for filtering the output of said first time reversal buffer in a second low pass filter; and

means for storing the output of said second low pass filter in a second time reversal buffer;

means for storing said tone correction function in a lookup table memory;

means for applying said stored plurality of pixels of said image plane in said memory to said lookup table memory to provide a plurality of tone corrected image output pixels from said lookup table memory corresponding to a tone corrected image; and

means for displaying said tone corrected image on a display device.

15. In an automatic tone correction system wherein a stored image plane including a plurality of pixels, is processed by applying a tone correction function to each pixel of said stored image plane, an apparatus for generating a tone corrected image from said stored image, comprising:

means for compiling a histogram of the relative occurrence of tone values corresponding to said stored pixels of said stored image plane;

means for processing said histogram, said histogram processing comprising;

means for adding a constant to said histogram; and

means for processing said histogram with a non-linear function so that small histogram values have relatively more contribution to said tone correction function, and large histogram values have relatively less contribution to said tone correction function, said non-linear function comprising raising each histogram value to a power between 0 and 1; and

means for filtering said histogram to smooth out the peaks of said histogram;

means for numerically integrating said processed histogram;

means for normalizing said numerically integrated histogram to extend over the desired range of tone values in interest in order to form said tone correction function;

means for storing said tone correction function in a lookup table memory; and

means for applying said stored plurality of pixels of said stored image plane in said memory to said lookup table memory to provide a plurality of tone corrected image output pixels from said lookup table memory corresponding to said tone corrected image.

16. In an automatic tone correction system wherein a stored image plane including a plurality of pixels, is processed by applying a tone correction function to each pixel of said stored image plane, an apparatus for generating a tone corrected image from said stored image, comprising:

means for compiling a histogram of the relative occurrence of tone values corresponding to said stored pixels of said stored image plane;

means for processing said histogram, said histogram processing comprising;

means for adding a constant to said histogram;

means for processing said histogram with a non-linear function so that small histogram values have relatively more contribution to said tone correction function, and large histogram values have relatively less contribution to said tone correction function comprising means for computing the square root of each tone value; and

means for filtering said histogram to smooth out the peaks of said histogram;

means for integrating said processed histogram;

means for normalizing said numerically integrated histogram to extend over the desired range of tone values in interest in order to form said tone correction function;

means for storing said tone correction function in a lookup table memory; and

means for applying said stored plurality of pixels of said stored image plane in said memory to said lookup table memory to provide a plurality of tone corrected image output pixels from said lookup table memory corresponding to said tone corrected image.

17. In an automatic tone correction system wherein a color image is stored in the form of a plurality of color planes, said color image being processed by applying a tone correction function to each pixel of said stored color planes, an apparatus for generating a tone corrected image from said stored image, comprising:

means for scanning a photographic image to provide a plurality of pixels corresponding to each of said color planes;

means for storing said plurality of pixels corresponding to each of said color planes in a memory;

means for compiling a histogram of the relative occurrence of tone values corresponding to said stored pixels of each of said color planes of said color image;

means for averaging said plurality of histograms to form a composite histogram;

means for processing said composite histogram with a non-linear function so that small histogram values have relatively more contribution to said tone correction function, and large histogram values have relatively less contribution to said tone correction function, said non-linear function comprising raising each histogram value to a power between 0 and 1;

means for numerically integrating said processed histogram;

means for normalizing said numerically integrated histogram to extend over the desired range of tone values in interest in order to form said tone correction function;

means for storing said tone correction function in a lookup table memory; and

means for applying said stored plurality of pixels of one of said color planes in said memory to said lookup table memory to provide a plurality of tone corrected image output pixels from said lookup table memory corresponding to said tone corrected image.

18. In an automatic tone correction system wherein a color image is stored in the form of a plurality of color planes, said color image being processed by applying a tone correction function to each pixel of said stored color planes, an apparatus for generating a tone corrected image from said stored image, comprising:

means for compiling a histogram of the relative occurrence of tone values corresponding to said stored pixels of each of said color planes of said color image;

means for averaging said plurality of histograms to form a composite histogram;

means for processing said composite histogram with a non-linear function so that small histogram values have relatively more contribution to said tone correction function, and large histogram values have relatively less contribution to said tone correction function, wherein said processing means further comprises means for computing the square root of each tone value;

means for numerically integrating said processed histogram;

means for normalizing said numerically integrated histogram to extend over the desired range of tone values in interest in order to form said tone correction function;

means for storing said tone correction function in a lookup table memory; and

means for applying said stored plurality of pixels of said color planes in said memory to said lookup table memory to provide a plurality of tone corrected image output pixels from said lookup table memory corresponding to said tone corrected image.