ADAPTIVE CONTRAST OPTIMIZATION OF DIGITAL COLOR IMAGES

Abstract

Methods and systems are provided for adjusting the contrast and intensity of digital images. Digital images may be processed by first creating a histogram that identifies the intensity of components of the digital image. A transformation function is then computed using the histogram. The transformation function can be used to adjust the intensity of a low light portion of the digital image and is applied to individual color components of the digital image. Finally, the contrast of a portion of the digital image other than the low light portion of the digital image is adjusted.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

The foregoing summary of the invention, as well as the following detailed description of illustrative embodiments, is better understood when read in conjunction with the accompanying drawings, which are included by way of example, and not by way of limitation with regard to the claimed invention.

FIG. 1 illustrates an exemplary mobile terminal that may be used to implement aspects of the present invention.

FIG. 2 illustrates a method of processing digital images in accordance with an embodiment of the invention.

FIG. 3 illustrates an exemplary histogram that includes intensity data of an intensity image in accordance with an embodiment of the invention.

FIG. 4 illustrates an exemplary method that may be used to compute a transformation function in accordance with an embodiment of the invention.

FIG. 5 illustrates an exemplary transformation function in accordance with an embodiment of the invention.

FIG. 6A illustrates an image prior to undergoing the image processing herein.

FIG. 6B illustrates an image that has undergone processing in accordance with an embodiment of the invention.

DETAILED DESCRIPTION

In the following description of various illustrative embodiments, reference is made to the accompanying drawings, which form a part hereof, and in which is shown, by way of illustration, various embodiments in which the invention may be practiced. It is to be understood that other embodiments may be utilized and structural and functional modifications may be made without departing from the scope of the present invention.

Aspects of the invention may be implemented with digital cameras and devices that include digital cameras. Devices that include digital cameras include mobile terminals and portable computer devices. FIG. 1 illustrates an exemplary mobile terminal 10 that may be used to implement aspects of the invention. As shown in FIG. 1, a mobile terminal 10 may include a processor 128 connected to a user interface 130, a memory 134 and/or other storage, a display 136 and a camera 156. Mobile terminal 10 may also include a battery 150, a speaker 152 and antennas 154. User interface 130 may further include a keypad, touch screen, voice interface, one or more arrow keys, joy-stick, data glove, mouse, roller ball, touch screen, or the like.

Computer executable instructions and data used by processor 128 and other components within mobile terminal 10 may be stored in computer readable memory 134. Memory 134 may be implemented with any combination of read only memory modules or random access memory modules, optionally including both volatile and nonvolatile memory and optionally being detachable. Software 140 may be stored within memory 134 and/or storage to provide instructions to processor 128 for enabling mobile terminal 10 to perform various functions. Alternatively, some or all of mobile terminal 10 computer executable instructions may be embodied in hardware or firmware (not shown).

Mobile terminal 10 may be configured to send and receive transmissions based on the Bluetooth standard, through a specific Bluetooth module 141. Additionally, mobile terminal 10 may also be configured to receive, decode and process transmissions through an FM/AM radio receiver 142, a wireless local area network (WLAN) transceiver 143, and a telecommunications transceiver 144. In one aspect of the invention, mobile terminal 10 may receive radio data stream (RDS) messages. Mobile terminal 10 may be equipped with other receivers/transceivers, e.g., one or more of a Digital Audio Broadcasting (DAB) receiver, a Digital Radio Mondiale (DRM) receiver, a Forward Link Only (FLO) receiver, a Digital Multimedia Broadcasting (DMB) receiver, etc. Hardware may be combined to provide a single receiver that receives and interprets multiple formats and transmission standards, as desired. That is, each receiver in a mobile terminal device may share parts or subassemblies with one or more other receivers in the mobile terminal device, or each receiver may be an independent subassembly.

FIG. 2 illustrates a method of processing digital images in accordance with an embodiment of the invention. First, in step 202 and intensity image that identifies the intensity of pixels of an input image is computed. The input image may be in a compressed or uncompressed RGB format such as JPEG and TIFF. In some embodiments the input image may be in the form of unprocessed data received from an image sensor of a digital camera. The intensity image may be computed using the following formula:

$\begin{array}{cc}y\left(i,j\right)=\frac{r\left(i,j\right)+b\left(i,j\right)+g\left(i,j\right)}{3}& \left(1\right)\end{array}$

where i identifies a pixel row,j identifies a pixel column. The intensity of individual red, green and blue pixels are identified by r(i,j), g(i,j) and b(i,j) respectively.

Next, in step 204 a histogram of the intensity image is created. The type of histogram created in step 204 may be a conventional histogram used in connection with the processing of typical digital images. FIG. 3 illustrates an exemplary histogram 300 that includes intensity data of an intensity image. As is known in the art, the quantities of low light pixels are shown toward the left of the histogram and the quantities of bright pixels are shown toward the right. Histogram 300 corresponds to an image having a large number of low light or dark pixels and relatively few bright pixels.

In one embodiment, the histogram created in step 204 may be defined as follows:

h(m)=nr of pixels in the range [m−0.5;m+0.5] with m=0, . . . , 255   (2)

In step 206 a transformation function is computed to adjust the intensity of a low light portion of the digital image. The computation performed in step 206 may use the histogram created in step 204. FIG. 4 illustrates an exemplary method that may be used to compute the transformation function. First, in step 402 the maximum intensity value MAX in the low light portion of the histogram and its position M are estimated. In one embodiment:

$\begin{array}{cc}\mathrm{MAX}=\underset{m=1,\dots ,128}{\max }\left\{h\left(m\right)\right\}\mathrm{and}M=\{m=1,\dots ,128h\left(m\right)=\mathrm{MAX}\}& \left(3\right)\end{array}$

In this manner the maximum number of pixels that are in the low light portion of the image and the corresponding number of pixels are found. The position M in the histogram corresponds to the value MAX. The low light portion of the histogram is analyzed so that the intensity and contrast of dark regions of the image can be adjusted.

In step 404, the width of the low light portion of the histogram is estimated. Step 404 may be used to determine the portion of the low light region containing the most pixels. In one embodiment, a threshold value, such as 0.3, is selected and the range of the histogram [Imin,Imax] is computed as:

$\begin{array}{cc}\mathrm{low}=\{m=1,\dots ,Mh\left(m\right)>0.3\mathrm{MAX}\}\mathrm{and}R1=\underset{m}{\min }\left\{\mathrm{low}\right\}& \left(4\right)\\ \mathrm{high}=\{m=M,\dots ,128h\left(m\right)>0.3\mathrm{MAX}\}\mathrm{and}R2=\underset{m}{\max }\left\{\mathrm{high}\right\}& \left(5\right)\end{array}$ Imin=M−max{|M−R1|,|M−R2|}and Imax=M+max{|M−R1|,|M−R2|}  (6)

Input images that have a large number of dark pixels that have almost the same intensity level will result in a small range (Imin and Imax will be close to each other). The amount of processing that ultimately takes place is inversely related to the range. Input images having large ranges already have a large dynamic range and require less processing. In some embodiments the maximum and the minimum strength of the processing are limited in order to limit over-processing or smoothing the input image.

Finally, in step 406 parameters of the transformation function are estimated and the transformation function is fit to the histogram. The transformation function may be fitted to the range [Imin,Imax] determined above. In one embodiment, the transformation function f(i,j) is defined as

$\begin{array}{cc}f\left(i,j\right)=\frac{255}{1+\exp \left(-\frac{K(y\left(i,j\right)-M}{255}\right)}& \left(7\right)\end{array}$

A scaling factor of 255 is used to normalize the intensity pixel values in the range [−1,1]. FIG. 5 illustrates an exemplary transformation function 500. The maximum slope exists for inputs in the range [−1,1] and this is the reason why the pixel values must be scaled. Subtracting the maximum intensity position M shifts the center of the correction function to the position of the maximum of the histogram.

The value K in equation can result in saturation at the zero level or the 255 level if not chosen properly. A very large value of K for a transformation function enhance the contrast of the pixels that have values close to M but the other pixels will be simply put to 0 or 255. That is all pixels not located close to M will be black or white. K may have a non-fixed value when there is a desire for high contrast while avoiding over saturation. A wide range [Imin,Imax] may result in a small K that prevents saturation of the other pixels and also decrease the strength of the processing. A narrow range [Imin,Imax] may result in a larger K in order to better increase the contrast and intensity. However in any case the pixels that are at the extreme value of the intensity (around 0 and around 255 intensity levels) should not be saturated.

In one embodiment, to compute the value of the parameter K we start from the following condition imposed on the transformation function f(i,j):

ƒ(Imax)=0.7*255   (8)

This value ensures a remaining dynamic range of 0.3*255 for pixels that are situated around 0 levels and also around 255 level. Solving equation (8) for K results in the following formula for adaptively selecting the value of this parameter:

$\begin{array}{cc}K=\frac{-255\ln \left(\frac{1-0.7}{0.7}\right)}{\mathrm{Imax}-M}=\frac{-255\ln \left(0.43\right)}{\mathrm{Imax}-M}& \left(9\right)\end{array}$

The scaling factor 255 ensures that the pixels after applying the correction function are on the range [0,255]. In some situations the value of the parameter K, estimated by equation (9), can be too large or to small. In such situations its value should be bounded. An exemplary minimum value Kmin equals 3 and an exemplary maximum value K max ε[8,10]. The fixed value Kmin may be chosen such that the transformation function f(i,j) is linear inside the input range [−1,1]. The maximum value of K can be selected fixed or it can be a user defined parameter. Modifications to Kmax can increase or decrease the maximum strength of the processing.

Returning to FIG. 2, in step 208 the transformation function is applied to individual color components of the digital image. In one implementation, for the red component r(i,j), the minimum and maximum values of r(i,j) are determined so that they may be preserved as follows:

$\begin{array}{cc}\mathrm{min\_r}=\underset{i,j}{\min }\left(r\left(i,j\right)\right)\mathrm{max\_r}=\underset{i,j}{\max }\left(r\left(i,j\right)\right)& \left(10\right)\end{array}$

Correction may then be performed as follows:

$\begin{array}{cc}\hat{r}\left(i,j\right)=\frac{\mathrm{max\_r}}{1+\exp \left(-\frac{K\left(r\left(i,j\right)-M\right)}{255}\right)}& \left(11\right)\end{array}$

The minimum and the maximum pixel values of the red color component may be restored as follows:

$\begin{array}{cc}\hat{r}\left(i,j\right)=\hat{r}\left(i,j\right)-\mathrm{min\_r}\hat{r}\left(i,j\right)=\frac{\mathrm{max\_r}-\mathrm{min\_r}}{\max \left(\hat{r}\left(i,j\right)\right)}\hat{r}\left(i,j\right)\hat{r}\left(i,j\right)=\hat{r}\left(i,j\right)+\mathrm{min\_r}& \left(12\right)\end{array}$

During the correction phase, some pixels from the high intensity part of the input image may be over processed. Over saturation can be corrected by using a weighted average, such as:

{circumflex over (r)}(i,j)=(1=a(i,j)){circumflex over (r)}(i,j)+a(i,j)r(i,j)   (13)

With a(i,j) being a weighting function computed as follows:

$\begin{array}{cc}a\left(i,j\right)=\{\begin{array}{cc}0& \mathrm{if}r\left(i,j\right)<M\\ \frac{f\left(y\left(i,j\right)\right)-0.5}{255}& \mathrm{otherwise}\end{array}& \left(14\right)\end{array}$

The processing described above with respect to red may then be repeated for other color components, such as green and blue.

Finally, in step 210 the contrast of a portion of the digital image other than the low light portion of the digital image is adjusted. In one embodiment, step 210 includes modifying the level of the local contrast, but not the intensity. The transformation function f(i,j) may be used to indicate which pixels were processed in the previous steps.

The contrast of a portion of the digital image other than the low light portion of the digital image may be adjusted by first computing a gain map g(i,j) according to:

$\begin{array}{cc}g\left(i,j\right)=\frac{1}{1+\exp \left(-\frac{K\left(r\left(i,j\right)-M\right)}{255}\right)}g\left(i,j\right)=g\left(i,j\right)-\min \left(g\left(i,j\right)\right)g\left(i,j\right)=\frac{g\left(i,j\right)}{\max \left(g\left(i,j\right)\right)}+1& \left(14\right)\end{array}$

In one implementation, a maximum value of the gain map is 2 and it is obtained for the pixels that were not processed at previous steps. Of course a maximum value larger than 2 can be used in order to have a stronger contrast processing. In an alternative embodiment, the maximum value of the gain can also be selected as a user defined parameter which can be used to fine tune the processing quality.

Vertical and horizontal sums and differences may be computed as follows:

d1={circumflex over (r)}(i,j)−{circumflex over (r)}(i+1,j)

s1={circumflex over (r)}(i,j)+{circumflex over (r)}(i+1,j)

d2={circumflex over (r)}(i,j)−{circumflex over (r)}(i,j+1)

s1={circumflex over (r)}(i,j)+{circumflex over (r)}(i,j+1)   (15)

Next, the portion of the digital image other than the low light portion of the digital image may be adjusted by multiplication of the differences with the gain map as follows:

{circumflex over (d)}1(i,j)=g(i,j)d1(i,j)

{circumflex over (d)}2(i,j)=g(i,j)d2)(i,j)   (16)

The image may then be reconstructed according to:

$\begin{array}{cc}\mathrm{Rout}\left(i,j\right)=\frac{\hat{d}1\left(i,j\right)+\hat{d}2\left(i,j\right)+s1\left(i,j\right)+s2\left(i,j\right)}{4}& \left(17\right)\end{array}$

Finally, image clipping may be performed according to:

$\begin{array}{cc}\mathrm{Rout}\left(i,j\right)=\{\begin{array}{cc}0& \mathrm{if}\mathrm{Rout}\left(i,j\right)<0\\ 255& \mathrm{if}\mathrm{Rout}\left(i,j\right)>255\\ \mathrm{Rout}\left(i,j\right)& \mathrm{otherwise}\end{array}& \left(18\right)\end{array}$

The remaining color components may be processed using similar procedures.

FIG. 6A illustrates an image 600 prior to undergoing the image processing described above. FIG. 6B illustrates an image 602 that has undergone processing in accordance with an embodiment of the invention. One can see that the intensity of the low light portion of image 602 exceeds the intensity of the low light portion of image 600.

While illustrative systems and methods as described herein embodying various aspects of the present invention are shown, it will be understood by those skilled in the art, that the invention is not limited to these embodiments. Modifications may be made by those skilled in the art, particularly in light of the foregoing teachings. For example, each of the elements of the embodiments may be utilized alone or in combination or subcombination with elements of the other embodiments. It will also be appreciated and understood that modifications may be made without departing from the true spirit and scope of the present invention. The description is thus to be regarded as illustrative instead of restrictive on the present invention.

Claims

1. A method of processing a digital image, the method comprising: (a) computing an intensity image that identifies the intensity of pixels of an input image;(b) creating a histogram of the intensity image;(c) computing a transformation function, using the histogram, that can be used to adjust the intensity of a low light portion of the digital image; and(d) applying the transformation function to individual color components of the digital image.

2. The method of claim 1, wherein (a) comprises determining the intensity of red, green and blue pixels.

3. The method of claim 2, wherein (a) comprises computing an intensity image y(i,j), such that

4. The method of claim 3, wherein the histogram of the intensity image y(i,j) is defined such that h(m)=nr of pixels in the range [m−0.5;m+0.5]with m=0, . . . , 255.

5. The method of claim 1, wherein the individual color components comprise red, green and blue color components.

6. The method of claim 1, wherein (c) comprises: (i) estimating a maximum number of pixels in a low light portion of the histogram;(ii) estimating a width of the low light portion of the histogram by determining a range that has intensity values that exceed a minimum intensity threshold; and(iii) computing the transformation function to apply to the low light portion of the digital image.

7. The method of claim 1, wherein (d) further comprises preserving minimum and maximum intensity values of each of the individual color components.

8. The method of claim 1, further including: (e) adjusting the contrast of a portion of the digital image other than the low light portion of the digital image.

9. The method of claim 1, wherein the adjustment in (e) is a function of the transformation function.

10. An apparatus comprising: a sensor that captures digital images;a processor programmed with computer-executable instructions to process a captured digital image by performing the steps comprising: (a) computing an intensity image that identifies the intensity of pixels of the captured digital image;(b) creating a histogram of the intensity image;(c) computing a transformation function, using the histogram, that can be used to adjust the intensity of a low light portion of the captured digital image; and(d) applying the transformation function to individual color components of the captured digital image.

11. The apparatus of claim 10, wherein the apparatus comprises a digital camera.

12. The apparatus of claim 10, wherein the apparatus comprises a mobile terminal.

13. The apparatus of claim 10, wherein (a) comprises computing an intensity image y(i,j), such that

14. The apparatus of claim 10, wherein the histogram of the intensity image y(i,j) is defined such that h(m)=nr of pixels in the range [m−0.5;m+0.5] with m=0, . . . , 255.

15. The apparatus of claim 10, wherein (c) comprises: (i) estimating a maximum number of pixels in a low light portion of the histogram;(ii) estimating a width of the low light portion of the histogram by determining a range that has intensity values that exceed a minimum intensity threshold; and(iii) computing the transformation function to apply to the low light portion of the captured digital image.

16. The apparatus of claim 10, wherein the processor is further programmed with computer-executable instructions to process a captured digital image by performing the step comprising: (e) adjusting the contrast of a portion of the captured digital image other than the low light portion of the captured digital image.

17. A computer-readable medium containing computer-executable instructions for processing a digital image by performing the steps comprising: (a) creating a histogram that identifies the intensity of components of a digital image;(b) computing a transformation function, using the histogram, that can be used to adjust the intensity of a low light portion of the digital image;(c) applying the transformation function to individual color components of the digital image; and(d) adjusting the contrast of a portion of the digital image other than the low light portion of the digital image.

18. The computer-readable medium of claim 17, wherein (b) comprises (i) estimating a maximum number of pixels in a low light portion of the histogram;(ii) estimating a width of the low light portion of the histogram by determining a range that has intensity values that exceed a minimum intensity threshold; and(iii) computing the transformation function to apply to the low light portion of the captured digital image.

19. An apparatus comprising: a means for capturing a digital image; anda means for adjusting intensity of a low light portion of a digital image and contrast of a portion of the digital image other than the low light portion of the digital image.

20. The apparatus of claim 19, further including: a means for sending and receiving data.