GB2456745A

GB2456745A - Luminance gamma correction with chroma gain adjustment

Info

Publication number: GB2456745A
Application number: GB0719632A
Authority: GB
Inventors: Christopher Paul Robinson; John George Cormack; Nicki John Paris
Original assignee: Keymed Medical and Industrial Equipment Ltd
Current assignee: Keymed Medical and Industrial Equipment Ltd
Priority date: 2007-10-08
Filing date: 2007-10-08
Publication date: 2009-07-29
Also published as: GB0719632D0

Abstract

A method of processing electronic image data in particular from a high-speed video camera comprises applying gamma correction 70 to the luminance component of a received image which is in luminance and colour difference (YCrCb) format, rather than first converting the luminance/chrominance image to red/green/blue format. Preferably the chroma components of the image are simultaneously subjected to a gain stage 80 which boosts the amplitude thereof, to compensate for the otherwise weakened colours in the gamma corrected image. The amount of gain applied to the colour difference (chroma) components, may be linked to the amount of applied gamma correction via a look up table or other relationship.

Description

Method and apparatus for image signal processing

Field of the invention

This invention relates to a method and an apparatus for processing an image signal and more particularly but not exclusively to gamma correction of high speed video images.

Background of the invention

Most solid-state electronic camera sensors have a linear response to light. That is, if the light falling on the sensor doubles in intensity, the output video signal also doubles in amplitude.

Many displays however, do not have a linear response.

In regions of video near to the black level, a small increase in input will result in a very small increase in brightness. In regions of video near to the white level, a small increase in input will result in a large increase in brightness.

This effect is well-known and is referred to as "Gamma", because the response of the display is mathematically represented as Light = Input where Y represents the Greek letter Gamma.

Many electronic cameras incorporate a "Gamma Corrector" (sometimes called an "Anti-Gamma") circuit which applies the inverse of the Gamma equation above to the electronic signal before it leaves the camera. Provided the value of gamma in the display and anti-gamma in the camera are the same, the resultant system will have a linear response to light.

In colour cameras, it is customary to apply gamma correction to the picture in RGB space. This means that anti-gamma is applied to the image when it is represented by the tristirnulus signals red, green and blue. The disadvantage of this technique is that in some cases the image must be converted into RGB from some other representation, for example luminance and colour difference signals such as YCrCb (see for example JP-A-2007 041296).

This incurs component cost and can reduce image quality through conversion loss and during signal processing, for

example.

In some cases, it is beneficial to apply more anti-gamma than is required to match the display's gamma curve, because this tends to brighten darker areas of the image without saturating brighter areas. The disadvantage of applying an excess of anti-gamma is that the colour saturation of the resultant image is reduced, since the red, green and blue signals tend to be compressed together in the brighter areas of the image. In other words, applying an excess of anti-gamma increases the image brightness somewhat but at the cost of reduced colour saturation. This effect is better understood with reference to Figures la and lb which show Red, Green and Blue output signals as a function of input amplitude, without and with gamma correction (that is, in accordance with a typical prior art gamma correction technique), respectively. It will be noted that the effect of gamma correction on the RGB signals is to compress them together towards the top right of the Figure (ib) where brightness is highest.

Summary of the Invention

Against this background, and in accordance with a first aspect of the present invention, there is provided a method of processing electronic image data comprising: receiving, at a signal processing means, an electronic signal containing data representative of an image, the image data defining colours within the image in terms of a luminance component (Y) and a plurality of colour difference chrominance (chroma) components (Cr, Cb) applying, at the signal processing means, gamma correction to the luminance component of the received image; and outputting, from the signal processing means, image output data including a gamma corrected luminance component.

The method of the present invention thus applies gamma correction directly to the luminance component Y which is typically the format in which video image data is obtained in, for example, a high speed video camera. This avoids the need to convert the image data (typically in the form of luminance Y and chrominance (chroma) colour difference data CrCb) to say, RGB format, reducing processor overhead, time and cost. Moreover, applying gamma correction to the luminance component allows for the application of excess anti-gamma without the disadvantageous loss of an excessive amount of colour saturation.

Although it is perfectly feasible and advantageous to apply anti gamma to the luminance component of the image data without any processing of the chroma components (i.e. the chrorna component either bypass the processor or pass through it substantially unaltered, in a particularly preferred embodiment, whilst the luminance component is being processed to apply anti gamma, the chroma components are subjected to a gain stage which boosts the amplitude thereof.

Boosting the chroma components when the luminance component is gamma corrected compensates for the otherwise weakening of the colours in the gamma corrected image data (anti gamma increases the amount of white in an image, which "washes out" colours).

In a further particularly preferred embodiment, the amount of gain applied to the colour difference (chroma) components is linked via, for example, a look up table or other relationship (defined empirically or otherwise) to the degree/amount of applied gamma correction. Thus a user may select a particular amount of gamma correction to apply to the luminance component, and this will result in a predetermined adjustment (such as a gain or boost) to the chroma components in consequence. This allows for a straightforward correction (by an amount previously determined to be appropriate, for example) to the otherwise reduced colour saturation resulting from the increased amount of white in the gamma corrected luminance components.

In accordance with a second aspect of the present invention there is provided signal processing apparatus for processing electronic image data comprising: an input, for receiving an electronic signal containing data representative of an image, the image data defining colours within the image in terms of a luminance component (Y) and a plurality of colour difference chroma components (Cr, Cb) ; a processor configured to receive the input image data, and to apply gamma correction to the luminance component thereof; and an output which provides an output signal including image data having a gamma corrected luminance component.

The invention also extends to a video camera, such as a high speed video camera, including such signal processing apparatus. -5-.

Throughout this specification, the term "luminance" and its representation "Y" are employed. Strictly, luminance is an optical physics term but the term "luma" (or 1') that is sometimes employed in image processing is usually in connection with the monochromatic component of an RGB signal to which gamma correction has been applied (see eg S<PTE EG28). Clearly in the present case such a term would therefore be inappropriate since the present invention specifically avoids the need to gamma correct an RGB signal.

In the foregoing and following, therefore, "luminance" or "Y" is intended to mean simply the monochromatic component of an image derived (or notionally derived) from linear RGB signals.

Further desirable features and advantages will be apparent from the appended claims.

Brief description of the drawings

The invention may be put in to practice in a number of ways, and a specific embodiment will now be described by way of example only and with reference to the accompanying figures in which: Figure la and Figure lb show, respectively, the red, green and blue components of an output signal as a function of input light, without and with gamma correction; Figure 2 shows a high schematic block diagram of the relevant components of a high speed video camera, including a signal processor for gamma correction of a high speed video signal; and Figure 3 shows the luminance component of an output signal of a camera as a function of the input light to the camera, with and without gamma correction respectively.

Detailed Description of a preferred embodiment

Referring first to Figure 2, a highly schematic block diagram of a high speed video camera 10 is shown. Light inputs the camera 10 through a lens 20 and is incident upon a CMOS image sensor30. The output of the CMOS image sensor is a digital signal containing digital video image data in the form, preferably, of luminance (Y) and colour difference (chrominance) components Cr and Cb. These data are subjected to pre-processing by pre-processor 40. The initial processing of the digital video image data by the pre-processor 40 typically includes analogue image capture, amplification, A to D conversion, storage and image processing/filtering (FPN removal, Bayer demosaicing, etc), amongst other processes. Many of these processes may be carried out via a digital signal processor (DSP) but the specific details of the preprocessor 40 do not form a part of this present invention and will therefore not be described in further detail. It is, however, to be observed that the pre-processor 40 in the arrangement of Figure 2 does not convert the representation of the input image data in YCrCb format into another format, such as RGB format, before the data is passed to the next stage in the camera 10.

The output of the pre-processor 40 is supplied to a gamma processing arrangement which is indicated generally at in Figure 2. The gamma processing arrangement 50 comprises a controller 60, whose purpose and manner of operation will be described in further detail below. The controller is in communication with an anti-gamma processor 70 for applying anti-gamma to a luminance component of an input signal from the pre-processor 40. The gamma processing arrangement 50 also contains a variable gain amplifier 80 which is capable of applying gain (boost) to the colour difference components (chroma components)Cr, Cb of the video image signal received from the pre-processor 40.

The controller 60 is also in communication with that variable gain amplifier 80.

The outputs of the anti gamma processor 70 and the variable gain amplifier 80 are post processed by post processor 90, to permit user adjustable brightness, contrast, colour saturation and inversion, dithering and so forth. Again the details of the post processor 90 do not form a part of the present invention. The output of the post processor 90 is provided as a digital or analogue video output video signal.

Although the gamma processing arrangements shown as a set of discrete block components in Figure 2 for the sake of ease of explanation, in the preferred embodiment the gamma processing circuitry 50 is merely a small part of a much larger colour processing arrangement which is partly or, most preferably, all implemented in firmware or software. For example, both the colour preprocessing (as briefly mentioned in the pre-processor 40 above) and the gamma correction may be implemented as software routines written in VHDL code and loaded into a DSP-FPGA. Thus it will be understood that the reference to separate components (for example), separate pre-processing 40, and/or separate anti-gamma and gain 70, 80) is merely for the sake of ease of explanation and that in the preferred embodiment there is or may not be any clear physical distinction between these components.

The manner of operation of the gamma processing 50 will now be described with reference to Figures 2 and 3. The anti-gamma processor 70 receives the luminance component of the digital video image data and applies anti-gamma to that luminance component. This allows more anti-gamma than is required to match the gamma curve of an output display, so that darker areas of the image may be brightened without saturating brighter areas. On the other hand, by applying the excess of anti-gamma to the luminance component only, the colour saturation of the resultant image is not excessively affected.

Although, for the above reasons, it is possible and advantageous simply to apply gamma correction to the luminance component whilst leaving the chrominance (colour difference) components unaffected by the gamma processing circuitry 50, in a particularly preferred embodiment, the chroma components Cr and Cb are supplied to the variable gain amplifier 80 as the luminance component Y is supplied to the anti-gamma processor 70. The reason for this may be understood by reference to Figure 3. Figure 3 shows the luminance component Y with and without gamma correction respectively. As may be seen in Figure 3, when anti-gamma is applied, the luminance component is increased relative to that luminance component without gamma correction. Without application of any change to the chroma components, the luminance becomes brighter with gamma correction but the colours are unchanged, which gives a weakening of the colours since they are "washed out" by the increase of white.

As explained above, in the preferred embodiment, the variable amplifier 80 (in particular) is embodied as VHDL code on an FPGA. Thus gain to the chroma components CrCb is applied as a simple multiplier (with + or -), the multiplication factor being obtained from a register loaded with a value sent along the parallel bus from the controller (microprocessor) 60.

To address this, the variable gain amplifier 80 to which the chroma components are applied provides a gain or boost to these chroma components. This allows for a compensation for the otherwise weakening of the colours through the gamma correction of the luminance components.

The controller 60 and the gamma processing arrangement are configured to control the anti gamma processor 70 and the variable gain amplifier 80 together so that, in the preferred embodiment of Figure 2, the amount of gain applied to the chroma components by the variable gain amplifier 80 is locked to the amount of anti-gamma applied to the luminance component. This may be achieved through, for example, the use of a look-up table or more sophisticated algorithm, based upon empirical observations of the way in which the human eye perceives images with different amounts of applied gamma correction. In other words, a user is able to define the amount of gamma correction to be applied through a user input 100 to the controller 60 of the gamma processing arrangement 50 in the camera 10. Application of the chosen level of anti-gamma to be applied by the anti gamma processor 70 to the luminance component may then automatically fix the amount of gain to be applied to the chroma components so as to provide an optimised, natural looking image. The following table sets out one typical example of the relative luminance gamma and chroma gain values that might be employed though it is stressed that these are merely exemplary; 11.00 1.00 [95 1.05 -10 -0.90 1.10 0.85 1.15 0.80 1.20 0.75 1.25 0.70 1.30 0.65 1.35 0.60 1.38 0.55 1.42 0.50 1. 45 0.45 1.46 0.40 1.47 0.35 1.47 0.30 1.47 In the preferred embodiment, the gamma correction is driven by calculations carried out in the controller (which is the system microprocessor, that communicates with the FPGA that provides the anti gamma via a parallel bus communication with registers in the latter). In other words, user input to the controller prompting a new amount of gamma results in calculations being carried out in the controller 60, and values then being sent via the bus to registers in the FPGA (anti gamma processor 70), for application to the video signals.

Although a preferred embodiment has been described, it will be appreciated that various modifications may be made without departing from the scope of the invention as defined by the appended claims. For example, although the camera of Figure 1 shows a CMOS image sensor 30, the skilled person would understand that any other image sensor such as, for example, a charge coupled device (CCD) would be suitable.

Moreover, the foregoing description is based upon the -11 -assumption that the signals being processed are digital video signals (hence the use of the digital colour difference terms CrCb) but it is to be understood that the invention is equally applicable to analogue signals (where the analogous format is sometimes referred to as YPrPb).

Indeed, the invention is applicable to any digital or analogue colour difference signal which has been derived from linear (not previously gamma processed) colour signals.

Claims

-12 -Claims: 1. A method of processing electronic data comprising: receiving, at a signal processing means, an electronic signal containing data representative of an image, the image data defining colours within the image in terms of a luminance component (Y) and a plurality of colour difference chrominance (chroma) components (Cr, Cb); applying, at the signal processing means, gamma correction to the luminance component of the received image; and outputting, from the signal processing means, imaging output data including a gamma corrected luminance component.
2. The method of claim 1, further comprising: passing the chroma components of the received image through the signal processing means without applying gamma correction thereto so that the output of the signal processing means includes image output data with a gamma corrected luminance component but with gamma uncorrected chroma components.
3. The method of claim 1 or claim 2, further comprising: applying a change, at the signal processing means, to the chroma components of the received image; and outputting from the signal processing means, image output data including both the gamma corrected luminance component and the changed chroma components.
4. The method of claim 3, wherein the step of applying a change to the chroma components comprises increasing the amplitude of the chroma components relative to that of the received image.

-13 -
5. The method of any preceding claim, further comprising adjusting the gamma correction which is applied to the received electronic signal.
6. The method of claim 5, wherein adjusting the gamma correction comprises altering the value of gamma representing the exponent in the relationship between the luminance Y and the image brightness.
7. The method of claim 4, further comprising adjusting the gamma correction which is applied to the luminance of the image data in the received electronic signal, and also applying the gain to the amplitude of the chroma components, wherein the degree of gamma correction is linked to the increase in amplitude of the gain applied to the chroma components such that, for any given level of anti-gamma applied to the luminance component, a predetermined gain is applied to the chrominance components.
8. The method of any preceding claim, wherein the electronic image data comprises video image data.
9. The method of claim 8, wherein the electronic image data comprises digital high speed video image data.
10. Signal processing apparatus for processing electronic image data comprising: an input, for receiving an electronic signal containing data representative of an image, the image data defining colours within the image in terms of a luminance component (Y) and a plurality of colour difference chroma components (Cr; Cb) ; a processor configured to receive the image data, and to apply gamma correction to the luminance component thereof; and an output which provides an output signal including image data having a gamma corrected luminance component.

-14 -
11. The apparatus of claim 10, further configured to pass the chroma components of the received image data to the output without the processor applying gamma correction thereto.
12. The apparatus of claim 11, wherein the processor is further configured to receive the input image data and to apply a change to the chroma components thereof so that the output of the signal processing apparatus includes both the gamma corrected luminance component and the changed chroma components.
13. The apparatus of claim 12, wherein the processor is configured to apply a gain to the chroma components of the input image data.
14. The apparatus of claim 12 or claim 13, wherein the processor includes a controller arranged (a) to receive an input from a user which defines the amount of gamma correction to be applied to the luminance component of the received image data; and (b) to control the processor to apply the user defined amount of gamma correction to the said received image data and to apply a predetermined change to the chroma components thereof which is related to the said user defined amount of gamma correction.
15. A high speed video camera including the signal processing apparatus of any of Claims 10 to 14.
16. A method of processing electronic image data substantially as herein described with reference to the accompanying figures.
17. A signal processing apparatus substantially as herein described with reference to the accompanying figures.

305S50; AJF; KC