CN103747225B - Based on the high dynamic range images double-screen display method of color space conversion - Google Patents

Abstract

The present invention relates to a kind of high dynamic range images double-screen display method based on color space conversion, it is characterized in that, comprise the following steps: S1: high dynamic range images is calculated and converts rgb space image to; S2: the rgb space image in step S01 is converted to HSL spatial image, completes being separated of brightness and tone; S3: the brightness L for the HSL spatial image in step S02 carries out independent self adaptation logarithmic mapping, completes the compression process of brightness, strengthens saturation S meanwhile; S4: the HSL spatial image in step S03 is converted to rgb space image; S5: front and back panel Iamge Segmentation is carried out to the rgb space image in step S04; S6: utilize the front and back panel image generated in step S05 to carry out double screen segmentation display in LCD-FED Double-Screen Display System.The present invention not only solves colour cast problem, and there has also been very large lifting on details expressive ability.

Description

Based on the high dynamic range images double-screen display method of color space conversion

Technical field

The present invention relates to the shuangping san field of high-dynamics image process and high dynamic range images, especially a kind of high dynamic range images double-screen display method based on color space conversion.

Background technology

In the last few years, the research for high dynamic range images got more and more, high dynamic range images be a kind of can the image of rediscover scene brightness scope.Its brightness range is from black night evening to sunlight fierce daytime, and dynamic range is even more than 1010:1.Compared to the brightness on a large scale of high dynamic range images, the dynamic range of current main-stream display device is just very little, nowadays the dynamic range of main flow display device is 400:1 to 600:1, the dynamic range of part display device can reach 1500:1, even if therewith, remain the requirement not reaching display high dynamic range images far away.For the solution of this problem, be carry out tone mapping on the one hand, these algorithms are roughly divided into two large classes: global map algorithm and local mapping algorithm.High dynamic range images can map and be compressed into low dynamic range echograms by tone-mapping algorithm, and for general display devices.No matter be global map algorithm, or local mapping algorithm, but these algorithms all can lose the part details of image, and all design for the object of the level visual rendition of the high dynamic range images of achromaticity (gray scale), after not considering that image brightness scope is compressed, colour gamut is also compressed, and causes the color of image visually to change, and produces colour cast problem.Be adopt shuangping san on the other hand, shuangping san can significantly improve the dynamic display capability of display system, can carry out details display better.

For colour cast problem, propose a lot of solution, but the MSRCR algorithm calculation of complex that people proposes such as these algorithms all have some shortcomings part, Rahma, and loss of detail is more; Algorithm based on histogram equalization has obvious distortion; Rapid brightness adjustment algorithm based on point shows not good enough in integral image harmony.

Existing shuangping san scheme comprises LCD-projector, LCD-LED and LCD-Optics.LCD-projector scheme is bulky, cannot be commonly public use.LCD-LED scheme but has that LED particle is comparatively large, easily produce the problems such as interference between pixel limits.The people such as GabrieleGuarnieri propose the Double liquid crystal display system of LCD-LCD structure, but this system brightness is lower, need high brightness backlights support, and refresh rate is also difficult to improve.The people such as PatrickLedda propose the Double-Screen Display System of LCD-Optics structure, but this systems bulky, limited viewing angle.

Summary of the invention

In view of this, the object of this invention is to provide a kind of high dynamic range images double-screen display method based on color space conversion, move the colour cast problem of range image process for height and details shows not enough problem.

The present invention adopts following scheme to realize: a kind of high dynamic range images double-screen display method based on color space conversion, is characterized in that, comprise the following steps:

S1: high dynamic range images is calculated and converts rgb space image to;

S2: the rgb space image in step S1 is converted to HSL spatial image, completes being separated of brightness and tone;

S3: the brightness L for the HSL spatial image in step S2 carries out independent self adaptation logarithmic mapping, completes the compression process of brightness, strengthens saturation S meanwhile;

S4: the HSL spatial image in step S3 is converted to rgb space image;

S5: front and back panel Iamge Segmentation is carried out to the rgb space image in step S4;

S6: utilize the front and back panel image generated in step S5 to carry out double screen segmentation display in LCD-FED Double-Screen Display System.

In an embodiment of the present invention, the concrete grammar of described step S1 is: if E is 0, then r=g=b=0.0, otherwise: r=R*2^ (E-128-8), g=G*2^ (E-128-8), b=B*2^ (E-128-8); Wherein (R, G, B, E) be RGBE image single pixel value, single pixel value that (r, g, b) is rgb space image.

In an embodiment of the present invention, described step S2 concrete grammar is: adopt sub-pixel corresponding conversion method by rgb space image value (r, g, b) HSL spatial image value (h, s, l) is converted to, complete image by the conversion of rgb space to HSL space, concrete switch process is as follows:

S21: rgb pixel value is transferred to the floating point values between (0,1);

S22: find out the maximum max in rgb pixel value and minimum value min, and calculate brightness $l=\frac{1}{2}(max+min);$

S23: if max=min, so S is defined as 0, and H is undefined and be defaulted as 0; Otherwise, calculate saturation S according to brightness L; Formula is: $s=\left\{\begin{array}{c}0,l=0ormax=min\\ \frac{\max -min}{2l},0<l\&le;\frac{1}{2}\\ \frac{\max -\min }{2-2l},l>\frac{1}{2}\end{array}\right.;$

S24: calculate tone H, formula is:

In an embodiment of the present invention, the self adaptation logarithmic mapping formula in described step S3 is:

$L_d(x,y)=\frac{L_{dmax}(x,y)\&times;0.01}{{\log }_{10}(L_{wmax}+1)}\&times;\frac{{\log }_{10}\left(L_w\right(x,y)+1)}{{\log }_{10}\&lsqb;2+\left({\left(\frac{L_w(x,y)}{L_{w\max }}\right)}^{\frac{{\log }_{10}\left(b\right)}{{\log }_{10}\left(0.5\right)}}\right)\&times;8\&rsqb;},$ Wherein, L _dmaxfor the maximum brightness value that display device can show; L _w(x, y) is the brightness value of each pixel; L _wmaxfor image brightness maximum; L _dthe brightness value that (x, y) is mapped image, parameter b is the overall brightness of adjustable parameters, effect diagram picture, and span is (0.5,1.0).

In an embodiment of the present invention, after the mapping calculation of described brightness L completes, carry out a non-linear gamma and correct, namely the power exponent computing that an index is 1/2.2 is carried out to brightness L.

In an embodiment of the present invention, enhancing formula carried out to saturation S be in described step S3: S'=α S, S' are the saturation after strengthening, and α is saturation enhancer.

In an embodiment of the present invention, described step S5 is specially: based on the thought of LCIS algorithm, and backpanel image is calculated by the square root of former figure, and frontpanel image is divided by by former figure and backpanel image and is calculated.

In an embodiment of the present invention, described backpanel image is used for the display of FED backlight, and described frontpanel image is used for LCD display.

The main feature of the inventive method is as follows:

High-dynamics image is carried out the conversion of color space, isolate color and brightness, process brightness value separately, we disregard the colouring information separated from high-dynamics image, just can effectively reduce colour cast degree.Again in conjunction with LCD-FED shuangping san, further increase the display capabilities of high-dynamics image details.Like this while image color can be retained, retain the details that image is abundanter.

Accompanying drawing explanation

Fig. 1 is method flow diagram of the present invention;

Fig. 2 a is high dynamic range original graph;

Fig. 2 b is HSL space original graph;

Fig. 3 a is HSL space L component map;

Fig. 3 b schemes (b=0.8) after L component self adaptation logarithmic mapping;

Fig. 4 a is direct RGB logarithmic mapping result figure;

Fig. 4 b is histogram equalization algorithm design sketch;

Fig. 4 c is HSL space transforming design sketch of the present invention;

Fig. 5 a is FED backlight display figure of the present invention;

Fig. 5 b is LCD display figure of the present invention;

Fig. 6 is the final shuangping san figure of the present invention.

Embodiment

For making object of the present invention, technical scheme and advantage clearly understand, below by specific embodiment and relevant drawings, the present invention will be described in further detail.

As shown in Figure 1, the invention provides a kind of high dynamic range images double-screen display method based on color space conversion, comprise the following steps:

S1: high dynamic range images RGBE (as shown in Figure 2 a) is calculated and converts rgb space image to;

S2: the rgb space image in step S1 is converted to HSL spatial image (as shown in Figure 2 b), completes being separated of brightness and tone;

S3: the brightness L for the HSL spatial image in step S2 carries out independent self adaptation logarithmic mapping, completes the compression process of brightness, strengthens accordingly saturation S meanwhile;

S4: the HSL spatial image in step S3 is converted to rgb space image; Namely reverse operating is carried out to step S2, HSL spatial image is rotated back into rgb space;

S5: front and back panel Iamge Segmentation is carried out to the rgb space image in step S4;

S6: utilize the front and back panel image generated in step S5 to carry out double screen segmentation display in LCD-FED Double-Screen Display System.

The expansion of RGBE file is called .hdr; So-called E, be exactly index, each passage of RGBE file is 8bit data type, and 4 passages come to 32bit; RGBE formatted file directly cannot carry out mapping process, need change.The concrete grammar of described step S1 is: if E is 0, then r=g=b=0.0, otherwise: r=R*2^ (E-128-8), g=G*2^ (E-128-8), b=B*2^ (E-128-8); Wherein (R, G, B, E) be RGBE image single pixel value, single pixel value that (r, g, b) is rgb space image.

Described step S2 concrete grammar is: adopt sub-pixel corresponding conversion method that rgb space image value (r, g, b) is converted to HSL spatial image value (h, s, l), complete image by the conversion of rgb space to HSL space, concrete switch process is as follows:

S21: rgb pixel value is transferred to the floating point values between (0,1);

S22: find out the maximum max in rgb pixel value and minimum value min, and calculate brightness $l=\frac{1}{2}(max+min);$

S23: if max=min, namely represent grey, so S is defined as 0, and H is undefined and be defaulted as 0; Otherwise, calculate saturation S according to brightness L; Formula is: $s=\left\{\begin{array}{c}0,l=0ormax=min\\ \frac{\max -min}{2l},0<l\&le;\frac{1}{2}\\ \frac{\max -\min }{2-2l},l>\frac{1}{2}\end{array}\right.;$

S24: calculate tone H, formula is: hue angle h ∈ [0,360), and h=0 presentation video is grey, and s, l are saturation and brightness, its span is all in [0,1].

Self adaptation logarithmic mapping formula in described step S3 is:

$L_d(x,y)=\frac{L_{dmax}(x,y)\&times;0.01}{{\log }_{10}(L_{wmax}+1)}\&times;\frac{{\log }_{10}\left(L_w\right(x,y)+1)}{{\log }_{10}\&lsqb;2+\left({\left(\frac{L_w(x,y)}{L_{w\max }}\right)}^{\frac{{\log }_{10}\left(b\right)}{{\log }_{10}\left(0.5\right)}}\right)\&times;8\&rsqb;},$

Wherein, L _dmaxfor the maximum brightness value that display device can show, as the quantization parameter of display, usually get L _dmax=100; L _w(x, y) is the brightness value of each pixel; L _wmaxfor image brightness maximum; L _dthe brightness value that (x, y) is mapped image.During logarithmic transformation, the selection of radix directly affects the quality for the treatment of effect, and radix regulates formula as follows: parameter b is the overall brightness of adjustable parameters, effect diagram picture, and span is generally (0.5,1.0).Experiment finds, best to the treatment effect of most of high-dynamics image during b=0.8.

After the mapping calculation of described brightness L completes, also need to carry out a non-linear gamma and correct, namely the power exponent computing that an index is 1/2.2 is carried out to brightness L.As best shown in figures 3 a and 3b, Fig. 3 is HSL space L component map, and Fig. 3 b schemes (b=0.8, gamma=1/2.2) after L component self adaptation logarithmic mapping.

Enhancing formula carried out to saturation S be in described step S3: S'=α S, S' are the saturation after strengthening, and α is saturation enhancer, general α=1.6.After brightness mapping and saturation strengthen, image is rotated back into rgb space by HSL space by again.As shown in Fig. 4 a, 4b and 4c, in Fig. 4 a, the image of direct RGB conversion differs larger with former figure on integral color, in Fig. 4 b, the performance of algorithm of histogram equalization color is better than Fig. 4 a, but red circle part has certain distortion, Fig. 4 c is the effect through HSL space transforming, image is not only very nearly the same in color with former figure, and there will not be problem of dtmf distortion DTMF.

In order to obtain better display effect, carry out more image detail display.The present invention adopts LCD-FED double screen high dynamic range display system to show processed image, shuangping san must carry out the front and back screen segmentation of input picture, existing double screen partitioning algorithm is LCIS algorithm, LCIS algorithm high dynamic range images is divided into basic brightness layer and levels of detail, but the hierarchical algorithm calculation of complex of LCIS, operand is large, and the efficiency of process image is very low.The present invention is based on the thought of LCIS (LowCurvatureImageSimplifier) algorithm, a kind of simple partitioning algorithm is adopted to carry out layering to image, improve image processing speed, rear board (backlight) image is calculated by the square root of former figure, and front panel (LCD) image is calculated by former figure and rear board (backlight) image subtraction.Specific algorithm formula is as follows: L _b(x, y)=F (L _in(x, y)), wherein, the value of image at (x, y) place is L _w(x, y), the brightness value of FED backlight is L _b(x, y), the brightness value of front panel LCD is L _f(x, y).Function F (L _in(x, y)) be a square root.As shown in figure 5a and 5b, Fig. 5 a is FED backlight display figure, Fig. 5 b is LCD display figure.Described rear board (backlight) image is used for the display of FED backlight, and described front panel (LCD) image is used for LCD display, as shown in Figure 6, finally obtains the image synthesized.

Above-listed preferred embodiment; the object, technical solutions and advantages of the present invention are further described; be understood that; the foregoing is only preferred embodiment of the present invention; not in order to limit the present invention; within the spirit and principles in the present invention all, any amendment done, equivalent replacement, improvement etc., all should be included within protection scope of the present invention.

Claims (8)

1., based on a high dynamic range images double-screen display method for color space conversion, it is characterized in that, comprise the following steps:

S1: high dynamic range images is calculated and converts rgb space image to;

S2: the rgb space image in step S1 is converted to HSL spatial image, completes being separated of brightness and tone;

S3: the brightness L for the HSL spatial image in step S2 carries out independent self adaptation logarithmic mapping, completes the compression process of brightness, strengthens saturation S meanwhile;

S4: the HSL spatial image in step S3 is converted to rgb space image;

S5: front and back panel Iamge Segmentation is carried out to the rgb space image in step S4;

S6: utilize the front and back panel image generated in step S5 to carry out double screen segmentation display in LCD-FED Double-Screen Display System.

2. the high dynamic range images double-screen display method based on color space conversion according to claim 1, it is characterized in that, the concrete grammar of described step S1 is: if E is 0, then r=g=b=0.0, otherwise: r=R*2^ (E-128-8), g=G*2^ (E-128-8), b=B*2^ (E-128-8); Wherein (R, G, B, E) be RGBE image single pixel value, single pixel value that (r, g, b) is rgb space image.

3. the high dynamic range images double-screen display method based on color space conversion according to claim 1, it is characterized in that, described step S2 concrete grammar is: adopt sub-pixel corresponding conversion method by rgb space image value (r, g, b) HSL spatial image value (h, s, l) is converted to, complete image by the conversion of rgb space to HSL space, concrete switch process is as follows:

S21: rgb pixel value is transferred to the floating point values between (0,1);

S22: find out the maximum max in rgb pixel value and minimum value min, and calculate brightness $l=\frac{1}{2}(max+min);$

S23: if max=min, so S is defined as 0, and H is undefined and be defaulted as 0; Otherwise, calculate saturation S according to brightness L; Formula is: $s=\left\{\begin{array}{c}0,l=0ormax=min\\ \frac{\max -min}{2l},0<l\&le;\frac{1}{2}\\ \frac{\max -\min }{2-2l},l>\frac{1}{2}\end{array}\right.;$

S24: calculate tone H, formula is:

4. the high dynamic range images double-screen display method based on color space conversion according to claim 1, is characterized in that: the self adaptation logarithmic mapping formula in described step S3 is:

$L_d(x,y)=\frac{L_{dmax}(x,y)\&times;0.01}{{\log }_{10}(L_{wmax}+1)}\&times;\frac{{\log }_{10}\left(L_w\right(x,y)+1)}{{\log }_{10}\&lsqb;2+\left({\left(\frac{L_w(x,y)}{L_{w\max }}\right)}^{\frac{{\log }_{10}\left(b\right)}{{\log }_{10}\left(0.5\right)}}\right)\&times;8\&rsqb;},$

Wherein, L _dmaxfor the maximum brightness value that display device can show; L _w(x, y) is the brightness value of each pixel; L _wmaxfor image brightness maximum; L _dthe brightness value that (x, y) is mapped image, parameter b is the overall brightness of adjustable parameters, effect diagram picture, and span is (0.5,1.0).

5. the high dynamic range images double-screen display method based on color space conversion according to claim 4, it is characterized in that: after the mapping calculation of described brightness L completes, carry out a non-linear gamma to correct, namely the power exponent computing that an index is 1/2.2 is carried out to brightness L.

6. the high dynamic range images double-screen display method based on color space conversion according to claim 1, it is characterized in that, enhancing formula carried out to saturation S be in described step S3: S'=α S, S' are the saturation after strengthening, and α is saturation enhancer.

7. the high dynamic range images double-screen display method based on color space conversion according to claim 1, it is characterized in that, described step S5 is specially: based on the thought of LCIS algorithm, backpanel image is calculated by the square root of former figure, and frontpanel image is divided by by former figure and backpanel image and is calculated.

8. the high dynamic range images double-screen display method based on color space conversion according to claim 7, is characterized in that: described backpanel image is used for the display of FED backlight, and described frontpanel image is used for LCD display.