Method and circuit arrangement for enhancing the contrast of an image

• US 6,873,742 B2
• Filed: 11/13/2001
• Issued: 03/29/2005
• Est. Priority Date: 06/15/1999
• Status: Expired due to Fees

• Current Assignee: Quarterhill Inc. (f/k/a Wi-LAN Inc.)
• Original Assignee: Infineon Technologies AG
• Inventors: Schu, Markus
• Primary Examiner(s): Mehta, Bhavesh M.
• Assistant Examiner(s): Kassa, Yosef
• Application Number:
  US10/054,296
• Publication Number:
  US 20020136464A1
• Time in Patent Office: 1,232 Days
• Field of Search: 382/218, 382/219, 382/252, 382/260, 382/263, 382/264, 382/274, 382/275, 358/3.26, 358/3.27, 358/520, 358/450, 358/463
• US Class Current: 382/274
• CPC Class Codes: H04N 1/4074 using histograms

Abstract

A method and apparatus for enhancing the contrast of an image, where pixels of the image having a brightness less than a mean brightness of the image are transformed in accordance with a first straight-line function, and pixels having a brightness greater than the mean brightness of the image are transformed in accordance with a second straight-line function. Variable slope factors of the two straight-line functions are determined by an analysis of the brightness distribution of the transformed output image. The pixels, thus transformed, produce the contrast-enhanced output image.

Claims

1. A method for enhancing the contrast of an image, the image being composed of a multiplicity of pixels each having a specific brightness value, the method comprising the steps of:

a) determining a mean brightness value of the individual pixels of the image;

b) processing the brightness values of those pixels of the image that have a brightness less than or equal to the mean brightness value, in accordance with a first transformation function;

c) processing the brightness values of those pixels of the image that have a brightness greater than the mean brightness value, in accordance with a second transformation function; and

d) outputting the pixels having the brightness values processed in steps b) and c), respectively, in the form of a transformed output image;

wherein a first straight-line function having a first variable slope factor is the first transformation function and the second transformation function is a second straight-line function having a second variable slope factor, the first slope factor of the first straight-line function and the second slope factor of the second straight-line function being determined by an analysis of the transformed output image.

2. The method as claimed in claim 1, wherein a function f_L(x_k)=x_m+(x_k−

x_m)*a_L is used as the first transformation function, and a function f_U(x_k)=x_m+(x_k−

x_m)*a_U is used as the second transformation function, where x_m denotes the mean brightness value, x_k denotes the respective brightness value to be processed, a_L denotes the first slope factor of the first transformation function f_L(x_k) and a_U denotes the second slope factor of the second transformation function f_U(x_k).

3. The method as claimed in claim 1, wherein the brightness values processed in accordance with the first transformation function are limited to a lower limit value when a brightness value less than the lower limit value results after the processing in accordance with the first transformation function, and the brightness values processed in accordance with the second transformation function are limited to an upper limit value when a brightness value greater than the upper limit value results after the processing in accordance with the first transformation function.

4. The method as claimed in claim 1, wherein the mean brightness value is determined by reducing the number of pixels of the image and calculating a quantized histogram for the image with the reduced pixels, from which histogram the mean brightness value is derived.

5. The method as claimed in claim 4, wherein the number of pixels of the image is reduced by low-pass filtering the image twice, each occurrence of filtering having an associated subsequent subsampling.

6. The method as claimed in claim 5, wherein the number of pixels is reduced by a factor 8 with each subsampling.

7. The method as claimed in claim 4, wherein the mean brightness value is calculated using the quantized histogram in accordance with the formula:

$x_m=\left(q_{\max }*\frac{L}{q}\right)+\frac{L}{2<br><br>q}$

where x_m denotes the mean brightness value, L denotes the number of possible brightness values, q denotes the number of quantization levels of the quantized histogram and q_max denotes the number of the quantization level having a maximum value of the quantized histogram, where 0≦

q_max≦

q−

1.

8. The method as claimed in claim 7, wherein when the quantized histogram has a plurality of quantization levels having the maximum value, the rounded-down mean value of the numbers of the quantization levels having the maximum value is used as a value for q_max.

9. The method as claimed in claim 4, wherein the quantized histogram has eight quantization levels for 256 possible brightness values.

10. The method as claimed in claim 4, wherein the step of determining the mean brightness value further comprises:

comparing the mean brightness value of the image with a mean brightness value of the preceding image and, dependent on a result of the comparison, an offset value is selected from a set comprising a plurality of discrete offset values and is added to the mean brightness value of the preceding image to achieve a resultant sum that is used as the mean brightness value for steps b) and c).

11. The method as claimed in claim 10, wherein the discrete offset values include −

1, 0 and 1.

12. The method as claimed in claim 1, wherein the first slope factor of the first straight-line function and the second slope factor of the second straight-line function are determined by an analysis of the brightness distribution of the transformed output image.

13. The method as claimed in claim 12, wherein a corresponding offset value is produced for each of the first and second slope factors as a function of the analysis of the brightness distribution of the transformed output image and is added to corresponding slope factors determined for the preceding image, in order to obtain the slope factors used for the processing of the instantaneous image in accordance with steps b) and c).

14. The method as claimed in claim 13, wherein one of a value −

OFFSET, 0, and +OFFSET is optionally used as the offset value, dependent on the analysis of the output image, where OFFSET denotes the magnitude of the offset value.

15. The method as claimed in claim 14, wherein OFFSET=1/128.

16. The method as claimed in claim 14, wherein during analysis of the brightness distribution of the transformed output image, a first number of pixels contained in the transformed output image having a brightness value less than or equal to a lower limit value is determined, a second number of pixels contained in the transformed output image having a brightness value greater than or equal to an upper limit value is determined, and both the first number of pixels and the second number of pixels are each compared with respective further upper and lower limit values in order to choose one of the values −

OFFSET, 0 and +OFFSET in each case depending on the comparison result for the offset values of the slope factors.

17. The method as claimed in claim 16, wherein the value −

OFFSET is chosen for the offset values of the slope factors of the first and second straight-line functions, respectively, when the first and second number of pixels respectively, are greater than the corresponding further upper limit values, the value +OFFSET is chosen for the offset values of the slope factor of the first and second straight-line functions, respectively, when the first and second number of pixels, respectively, are less than the corresponding further lower limit values and in all other cases the value 0 is chosen for offset values for the slope factors of the first and second straight-line function, respectively.

18. The method as claimed in claim 16, wherein a value of 16 is chosen as the lower limit value and a value of 240 is chosen as the upper limit value.

19. The method as claimed in claim 1, wherein the slope factors of the first and second straight-line functions are in each case limited to a maximum slope factor value of 1.5 and a minimum slope factor value of 1.

20. A circuit arrangement for enhancing the contrast of an image, wherein the image is comprised of a plurality of individual pixels each having a specific brightness value, the circuit arrangement comprising:

a mean value calculation unit configured to determine a mean brightness value of the individual pixels of the image;

a first transformation function calculation unit configured to determine a first transformation function that is used to process a first set of pixels of the image having a brightness less than or equal to the mean brightness value;

a second transformation function calculation unit configured to determine a second transformation function that is used to process a second set of pixels of the image having a brightness greater than the mean brightness value; and

an image processing unit configured for processing the brightness values of the first set of pixels of the image in accordance with the first transformation function and the brightness values of the second set of pixels of the image in accordance with the second transformation function and for outputting the processed pixels in the form of a transformed output image;

wherein the first and second transformation function calculation units are further configured to respectively determine a first straight-line function and a second straight-line function by analyzing the transformed output image;

the first and second straight-line functions respectively having a variable slope factor as the corresponding first and second transformation functions.

21. The circuit arrangement as claimed in claim 20, wherein:

the first transformation function calculation unit determines the function f_L(x_k)=x_m+(x_k−

x_m)*a_L as the first transformation function, and

the second transformation function calculation unit determines the function f_U(x_k)=x_m+(x_k−

x_m)*a_U as the second transformation function,

where x_m denotes the mean brightness value determined by the mean value calculation unit, x_k denotes the respective brightness value to be processed, a_L denotes the variable slope factor of the first transformation function f_L(x_k), and a_U denotes the variable slope factor of the second transformation function f_U(x_k).

22. The circuit arrangement as claimed in claim 20, wherein the mean value calculation unit further comprises:

at least one subsampling device in order to reduce the number of pixels of the image; and

a quantized histogram calculation unit that is configured to calculate a quantized histogram of the image having the reduced pixels and to determine the mean brightness value in accordance with the formula;

$x_m=\left(q_{\max }*\frac{L}{q}\right)+\frac{L}{2<br><br>q}$

where x_m denotes the mean brightness value, L denotes the number of possible brightness values, q denotes the number of quantization levels of the quantized histogram and q_max denotes the number of the quantization level having the maximum value of the quantized histogram, where 0≦

q_max≦

q−

1.

23. The circuit arrangement as claimed in claim 22, wherein when the quantized histogram has a plurality of quantization levels with the maximum value, the quantized histogram calculation unit uses a rounded-down mean value of the numbers of the quantization levels having the maximum value as a value for q_max.

24. The circuit arrangement as claimed in claim 22, wherein the mean value calculation unit further comprises:

an offset calculating unit that is configured to compare the mean brightness value of the image, the mean brightness value having been determined by the quantized histogram calculation unit, with a mean brightness value of a preceding image and select an offset value from a set comprising a plurality of discrete offset values and add the offset value to the value of the mean brightness of the preceding image depending on the comparison to achieve a sum, wherein the sum is fed as the mean brightness value to the first and second transformation function calculation units.

25. The circuit arrangement as claimed in claim 20, wherein:

the first transformation function calculation unit is assigned a pixel number determining device that is configured to determine a first number of pixels contained in the transformed output image, wherein the first number of pixels have a brightness value less than or equal to a lower limit value;

the second transformation function calculation unit is assigned the pixel number determining device that is further configured to determine a second number of pixels contained in the transformed output image, wherein the second number of pixels have a brightness value greater than or equal to an upper limit value;

the first and second transformation function calculation units are comprised of respective offset value determining units, which are configured to compare the first number of pixels and the second number of pixels, respectively, with corresponding further limit values in order to determine, dependent on results of the comparisons, first and second offset values, respectively, for the slope factor of the first and second straight-line functions, respectively; and

the first and second transformation function calculation units are each further comprised of a respective addition units configured for adding the first and second offset values, respectively, to the corresponding slope factors determined for the preceding image in order to obtain the respective slope factors of the first and second straight-line functions for the image that is to be presently processed.

26. The circuit arrangement as claimed in claim 25, wherein the offset value determining units are each configured to set the corresponding offset value to one of the values −

OFFSET, 0 and +OFFSET dependent on the result of the comparison.

27. The circuit arrangement as claimed in claim 26, wherein the OFFSET=1/128.

28. The circuit arrangement as claimed in claim 20, wherein the first and second transformation function calculation units each further comprise slope factor limiting units configured for limiting the determined slope factor of the first and second straight-line functions, respectively, to a maximum slope factor value of 1.5 and a minimum slope factor value of 1.