CN103024300A - Device and method for high dynamic range image display - Google Patents

Abstract

The invention is applicable to the technical field of image processing and provides a device and a method for high dynamic range image display. The method includes: segmenting a high dynamic range image according to preset image segmentation algorithm to acquire a plurality of image subblocks; adjusting the brightness value of any pixel point in each image subblock according to attribute information of each image subblock, the brightness value of the pixel point at the corresponding position of each image subblock and the preset image linear mapping relation; and generating and displaying a low dynamic range image according to the adjusted brightness value of each pixel point in each image subblock. Displaying effects of the generated low dynamic range image are greatly improved by adjusting the brightness value of any pixel point in each image subblock according to attribute information of each image subblock, the brightness value of the pixel point at the corresponding position of the high dynamic range image and the preset image linear mapping relation.

Description

A kind of high dynamic range images display packing and device

Technical field

The invention belongs to technical field of image processing, relate in particular to a kind of high dynamic range images display packing and device.

Background technology

Development along with the CMOS technology, camera in present most of mobile phone and network cameras are all built-in ISP (Image Signal Processing, the single chip module of the cmos sensor picture signal processing), all primary image Processing Algorithm such as automatic focusing, automatic exposure, Automatic white balance and BLC all have been embedded among the ISP.As everyone knows, the analog signal of imageing sensor output all is the above HDR(High Dynamic Range of 10bits through the image behind the analog-digital conversion a/d, high dynamic range) image, but existing display device, display screen such as the display of PC, digital television display, mobile phone, the dynamic range of images that can show only has the LDR(Low Dynamic Range of 8bits, low-dynamic range) image is lost so will have image information in the HDR image compression becomes the process of LDR image.How effectively the HDR image compression to be arrived the LDR image, also can be so that the display effect of image be better, this process is called TM(Tone Mapping, tone mapping).

At present more existing researchers are in the research of the high dynamic range images of paying close attention to real scene to the low dynamic range echograms compression.Existing tone mapping method is divided into two classes substantially: (1) global map, claim again the spatial domain invariant operator, and namely all pixels of image are used same compression curve, proofread and correct such as common image straight line drawing process, Gamma etc.; (2) local mapping claims again the spatial domain to change operator, and namely the mapping curve of each pixel is relevant with neighborhood territory pixel information.Although existing global map method can become the LDR image with the HDR image compression to a certain extent, avoid the impact of halation phenomenon, yet the method is based on empirical model, so that the LDR image reaches better effect, need to carry out careful adjusting etc. to relevant parameter; And existing local mapping method design is comparatively complicated, can not fundamentally solve the problem of halation impact etc.

Summary of the invention

The purpose of the embodiment of the invention is to provide a kind of high dynamic range images display packing and device, be intended to solve since utilize prior art with the HDR image compression for the LDR image after, the problem that the LDR image display effect is not good.

First aspect, described high dynamic range images display packing comprises:

According to default image segmentation algorithm, high dynamic range images is cut apart, obtain a plurality of image subblocks;

In the brightness value of described high dynamic range images corresponding position and the default linear mapping relations of image, adjust the brightness value of described arbitrary pixel in described each image subblock according to arbitrary pixel in the attribute information of each image subblock in described a plurality of image subblocks, described each image subblock;

According to the brightness value of each pixel in each image subblock after adjusting, generate and the demonstration low dynamic range echograms.

Second aspect, described high dynamic range images display unit comprises the first cutting unit, adjustment unit, image generation unit and display unit, wherein:

Described the first cutting unit is used for according to default image segmentation algorithm high dynamic range images being cut apart, and obtains and send a plurality of image subblocks to described adjustment unit;

Described adjustment unit, be used for receiving described a plurality of image subblocks that described the first cutting unit sends, reach the linear mapping relations of default image according to the brightness value of arbitrary pixel in the attribute information of each image subblock in described a plurality of image subblocks, described each image subblock in described high dynamic range images corresponding position, adjust the brightness value of described arbitrary pixel in described each image subblock, and the brightness value of each pixel is sent to described image generation unit in each image subblock after will adjusting;

Described image generation unit, brightness value for each each pixel of image subblock after the described adjustment that receives described adjustment unit transmission, and according to the brightness value of each pixel in each image subblock after adjusting, generate and send low dynamic range echograms to described display unit; And

Described display unit is used for receiving the described low dynamic range echograms that described image generation unit sends, and shows described low dynamic range echograms.

The embodiment of the invention is obtained the attribute information of a plurality of image subblocks of high dynamic range images by image segmentation, attribute information according to each image subblock, arbitrary pixel is in the brightness value of this high dynamic range images corresponding position and the default linear mapping relations of image in this each image subblock, adjust the brightness value of this arbitrary pixel in this each image subblock, so that according to the brightness value of each pixel in each image subblock after adjusting, effectively generate low dynamic range echograms, and method of operation is simple, the low dynamic range echograms that generates also has better display effect, solved since utilize prior art with the HDR image compression for the LDR image after, the problem that the LDR image display effect is not good.

Description of drawings

Fig. 1 is the realization flow figure of the high dynamic range images display packing that provides of first embodiment of the invention;

Fig. 2 is the realization flow figure of the high dynamic range images display packing that provides of second embodiment of the invention;

Fig. 3 (a), 3(b) be respectively display effect figure, the display effect figure after the Region Segmentation of the low dynamic range echograms that provides of the embodiment of the invention;

Fig. 4 is the structure chart of the high dynamic range images display unit that provides of third embodiment of the invention;

Fig. 5 is the structure chart of the high dynamic range images display unit that provides of fourth embodiment of the invention;

Fig. 6 is the structure chart of the high dynamic range images display unit that provides of fifth embodiment of the invention.

Embodiment

In order to make purpose of the present invention, technical scheme and advantage clearer, below in conjunction with drawings and Examples, the present invention is further elaborated.Should be appreciated that specific embodiment described herein only in order to explain the present invention, is not intended to limit the present invention.

The embodiment of the invention is by obtaining a plurality of image subblocks of high dynamic range images being cut apart rear generation, attribute information in conjunction with each image subblock in these a plurality of image subblocks, arbitrary pixel is in the brightness value of this high dynamic range images corresponding position and the default linear mapping relations of image in this each image subblock, adjust the brightness value of this arbitrary pixel in this each image subblock, again according to the brightness value of each pixel in each image subblock after adjusting, generate low dynamic range echograms, and this low dynamic range echograms has greatly improved the display effect of the low dynamic range echograms that generates more near this high dynamic range images.

Below in conjunction with specific embodiment specific implementation of the present invention is described in detail:

Embodiment one:

Fig. 1 shows the realization flow of the high dynamic range images display packing that first embodiment of the invention provides, and details are as follows:

In step S101, according to default image segmentation algorithm, high dynamic range images is cut apart, obtain a plurality of image subblocks.

In specific implementation process, the local display effect is better and Algorithm Performance is better in order to pursue, the embodiment of the invention mainly addresses the mapping effect aspect problem of high dynamic range images from these two aspects, by adopting the local mapping method of Region Segmentation, solve high/low brightness local in the image, the problem such as dizzy, and regulate local parameter in the operating process and can not affect other regional display effects.Particularly, can adopt existing image segmentation algorithm based on graph theory, should the scene image of complexity can be cut into the more consistent regional area of a plurality of internal features based on the image segmentation algorithm of graph theory, and after then cutting apart, can obtain a plurality of image subblocks of this high dynamic range images.

In step S102, in the brightness value of this high dynamic range images corresponding position and the default linear mapping relations of image, adjust the brightness value of this arbitrary pixel in this each image subblock according to arbitrary pixel in the attribute information of each image subblock in these a plurality of image subblocks, this each image subblock.

In step S103, according to the brightness value of each pixel in each image subblock after adjusting, generate and the demonstration low dynamic range echograms.

In specific implementation process, add up the attribute information of each image subblock in these a plurality of image subblocks, this attribute information comprises the maximum L of the pixel intensity of this each image subblock _Max, mean value L _Mean, variance L _StdDeng.Shown in the linear following formula of mapping relations of image (1) that should be default:

$L\left(p\right)=\frac{L\max }{1+\exp (-\mathrm{Lstd}\×(H\left(p\right)-\mathrm{Lmean}))},---\left(1\right)$

Wherein, H (p) represents that arbitrary pixel p is at the brightness value of this high dynamic range images corresponding position in this each image subblock, and L (p) represents the brightness value of this arbitrary pixel p in described each image subblock after the adjustment.

Suppose arbitrary image subblock Image1, corresponding attribute information comprises L1 _Max, L1 _Mean, L1 _StdAnd each pixel is at the brightness value of this high dynamic range images corresponding position among known this Image1, then the above-mentioned formula of substitution (1) can obtain the respective pixel brightness value of each pixel in this image subblock Image1, thereby the brightness value of each pixel among this Image1 has been carried out adjusting renewal.Equally, can adjust the brightness value of all pixels of each image subblock in these a plurality of image subblocks through above-mentioned steps.Then according to the brightness value of all pixels in all image subblocks after adjusting, can generate the low dynamic range echograms corresponding with this high dynamic range images, so that this low dynamic range echograms shows at display device.

In embodiments of the present invention, by at first high dynamic range images being carried out Region Segmentation, the different image subblock that gets access to is used respectively mapping relations or claims mapping curve to process, finally be converted to the image in the low-dynamic range, the problem of local display will more directly be processed, and it is also fairly simple to adjust mapping curve corresponding to each image subblock, can not affect the display effect of other area image sub-blocks in the image, the low state range image that gets access to the most at last or title LDR image are when display device shows, display effect is better.

Embodiment two:

Fig. 2 shows the realization flow of the high dynamic range images display packing that second embodiment of the invention provides, and details are as follows:

In step S201, high dynamic range images is compressed, obtain the low dynamic range echograms that display device can show.

In step S202, according to default image segmentation algorithm, this low dynamic range echograms is cut apart, obtain a plurality of image subblocks.

In specific implementation process, for high dynamic range images or the title HDR image of higher pixel, directly processing is more consuming time, in order to reduce the processing time, can carry out down-sampling to it first, be hanged down pixel or claim the image of low-dynamic range to carry out again relevant treatment afterwards.Such as, to with the HDR image mapped to the LDR image, can use the global map method, such as Linear Mapping method or S curve global map method first, to the LDR view data, specifically can use the 0-1 algorithm that the HDR image compression is become the LDR image HDR Image Data Compression.Then after being compressed into the LDR image, can use region segmentation method based on graph theory to the LDR Image Segmentation Using, and each image subblock of cutting apart is added up respectively corresponding attribute information, comprise pixel intensity maximum L _Max, mean value L _Mean, variance L _StdDeng.Fig. 3 (a) is depicted as this HDR image is carried out demonstration result after the 0-1 compression algorithm, and Fig. 3 (b) is for adopting based on the display effect figure after the region segmentation method of graph theory.

In step S203, in the brightness value of this high dynamic range images corresponding position and the default linear mapping relations of image, adjust the brightness value of this arbitrary pixel in this each image subblock according to arbitrary pixel in the attribute information of each image subblock in these a plurality of image subblocks, this each image subblock.

In step S204, according to the brightness value of each pixel in each image subblock after adjusting, generate and the demonstration low dynamic range echograms.

Wherein, the attribute information of each image subblock in these a plurality of image subblocks (or claiming mapping parameters) comprises the pixel intensity maximum L of this each image subblock _Max, mean value L _Mean, variance L _Std

As the present invention's one example, in conjunction with formula in above-described embodiment (1), by step S203, step S204, then can obtain the low dynamic range echograms after this high dynamic range images mapping, and can show at display device.

As another example of the present invention, because L _MeanValue has represented the average characteristics of image subblock brightness; L _StdRepresented pixel intensity and the degrees of offset of average brightness and the intensity of distribution of image subblock, and L _StdLess, be offset littlely, brightness is more concentrated.In order to obtain better image display effect, can get L according to practical experience _StdValue is 0.5; L _MaxIt is the maximum brightness value of image subblock, because Region Segmentation is not very accurate, so it is the characteristic of characterize target areas well, directly as mapping parameters and improper, and for darker target area, in order better its true effect to be shown, its overall brightness need to be promoted, in addition because the L that this moment the image subblock statistics is obtained _MaxUsually very little, during directly as mapping parameters, show that the result remains very dark, in actual mechanical process, can estimate L according to reliable relational expression _MaxValue, statistics is found, is worked as L _MeanValue hour, the L of use _MaxShould be larger, L _MeanBe worth when larger the L of use _MaxValue should be less, but L _MeanWith L _MaxIt is not strict inversely prroportional relationship.Then based on principle of least square method the result of a large amount of statistical analyses is carried out the linear polynomial match, obtained L _MeanWith L _MaxCorresponding relation, shown in the following formula (2):

$L_{\max }={10}^5\×\left(\begin{array}{c}0.266\×{\mathrm{Lmean}}^6-1.102\×{\mathrm{Lmean}}^5+1.476\×{\mathrm{Lmean}}^4\\ -0.829\×{\mathrm{Lmean}}^3+0.205\×{\mathrm{Lmean}}^2-0.021\×\mathrm{Lmean}+0.001\end{array}\right),---\left(2\right)$

Further, by the maximum L of the pixel intensity of each image subblock is set according to above-mentioned formula _Max, corresponding variance L is set _StdValue be 0.5, corresponding mean value L is set _MeanRemain unchanged, in conjunction with formula in above-described embodiment (1), by step S203, step S204, then can obtain the low dynamic range echograms after this high dynamic range images mapping, and can be more efficiently and accurately low dynamic range echograms after display device shows this mapping, also strengthened the user and experienced effect, user satisfaction is provided.

In embodiments of the present invention, because highly sensitive in colourity to brightness of human eye when luminance difference is larger between adjacent image sub-block or the zone, causes that easily human eye is uncomfortable.And in the actual displayed process, differ less adjacent image sub-block for average brightness, if adjust the L of latter two sub-block of pixel brightness value _MeanDifference is larger, then can produce false edge in procedure for displaying; And differ larger adjacent domain for average brightness, then in follow-up adjustment pixel brightness value process, should keep the edge that originally exists.For this problem, can further be optimized on the whole the low dynamic range echograms after the mapping.Also namely in this step S203, after this adjusted the brightness value of this arbitrary pixel in this each image subblock, before this step S204, this high dynamic range images display packing can also comprise:

Multiple continuous and nonoverlapping range of luminance values is set reaches one to one brightness degree;

According to the range of luminance values under the mean value of the pixel intensity of each image subblock, obtain brightness degree corresponding to this each image subblock;

According to the level spacing between brightness degree corresponding to adjacent image sub-block, adjust the maximum of the corresponding pixel intensity of image subblock that pixel brightness level is high in this adjacent image sub-block.

In specific implementation process, such as can but be not limited to arrange four kinds of continuous and nonoverlapping range of luminance values and four kinds of brightness degrees etc., shown in following table one:

Table one

Range of luminance values (unit: candela/metre2)	Brightness degree	Brightness level
			(-∞，0.1]	First	Darker
(0.1，0.8]	Second	Intermediate light
			(0.8，3]	The 3rd	Brighter
(3，+∞)	The 4th	Highlighted

Further, can carry out mark to these a plurality of image subblocks, be level-1 area such as the image subblock area maximum image subblock that number of pixels is maximum in other words is set, the nearest image subblock in center that whole image perhaps is set is level-1 area etc., after this All Ranges adjacent with this level-1 area being set is level-2 area, if the All Ranges adjacent with this level-1 area has a plurality of, then can continue to be labeled as level-2 area 1, level-2 area 2 etc., equally, according to this markedness theory, the adjacent area that continues the unmarked mistake of mark level-2 area is tertiary zone, by that analogy, until the complete all image subblocks of mark.

Such as, at first to some level-2 areas of this level-1 area, be optimized such as level-2 area 1, if the average brightness of this level-1 area belongs to (∞, 0.1] scope, corresponding the first brightness degree then, belong to darker zone, if corresponding the 3rd brightness degree of level-2 area 1 belongs to brighter zone, if perhaps level-2 area 1 corresponding the 4th brightness degree, belong to highlighted zone, can know that then the interval between brightness degree corresponding to this adjacent image sub-block differs at least one grade, then this moment, not need the image subblock of this level-2 area 1 is carried out brightness optimization; If level-2 area 1 corresponding the second brightness degree, the zone that belongs to intermediate light, if perhaps level-2 area 1 corresponding the first brightness degree, can know that then the interval between brightness degree corresponding to this adjacent image sub-block is more or less the same in a grade, then at this moment, continue to judge that whether the absolute value of difference of the average brightness that this adjacent image sub-block is corresponding is less than preset difference value, be, then do not need the image subblock of level-2 area 1 is carried out brightness optimization, no, then keep the average brightness L of the image subblock of level-2 area 1 _MeanAnd L _StdValue all constant, and upgrade corresponding L _MaxValue is the L of this adjacent level-1 area _MaxIdentical value.By that analogy, can optimize other adjacent level-2 areas of this level-1 area.Afterwards, continue to optimize the tertiary zone of the mark adjacent with level-2 area, until travel through all image subblocks.

In embodiments of the present invention, by high dynamic range images is compressed, obtain the low dynamic range echograms that display device can show, according to default image segmentation algorithm, this low dynamic range echograms is cut apart, obtain a plurality of image subblocks, again according to the attribute information of each image subblock in these a plurality of image subblocks, arbitrary pixel is in the brightness value of this high dynamic range images corresponding position and the default linear mapping relations of image in this each image subblock, realization is to the adjustment of the brightness value of this arbitrary pixel in this each image subblock, thereby according to the brightness value of each pixel in each image subblock after adjusting, can generate and show the low dynamic range echograms with better display effect.Further, can also be according to the brightness degree between the adjacent image sub-block, image to final demonstration is optimized, and state on the implementation in the Optimum Operation process, can also multiple continuous and nonoverlapping range of luminance values need be set according to reality and reach one to one brightness degree, brightness level etc., under the prerequisite that guarantees better display effect, also improved the flexibility ratio of this high dynamic range images display packing etc.

One of ordinary skill in the art will appreciate that all or part of step that realizes in above-described embodiment method is to come the relevant hardware of instruction to finish by program, described program can be stored in the computer read/write memory medium, described storage medium is such as ROM/RAM, disk, CD etc.

Embodiment three:

Fig. 4 shows the structure of the high dynamic range images display unit that third embodiment of the invention provides, and for convenience of explanation, only shows the part relevant with the embodiment of the invention.

This high dynamic range images display unit comprises processor 41 and display 42, wherein:

This processor 41, be used for according to default image segmentation algorithm, high dynamic range images is cut apart, obtain a plurality of image subblocks, according to arbitrary pixel in the attribute information of each image subblock in these a plurality of image subblocks, this each image subblock in the brightness value of this high dynamic range images corresponding position and the default linear mapping relations of image, adjust the brightness value of this arbitrary pixel in this each image subblock, brightness value according to each pixel in each image subblock after adjusting generates and sends low dynamic range echograms to this display 42; And

This display 42 is used for receiving this low dynamic range echograms that this processor 41 sends, and shows this low dynamic range echograms.

In embodiments of the present invention, because the image segmentation algorithm based on graph theory can be cut into the scene image of complexity the more consistent regional area of a plurality of internal features, segmentation effect is better, then can adopt this image segmentation algorithm based on graph theory that original high dynamic range images is cut apart by this processor 41, generate and the corresponding a plurality of image subblocks of this high dynamic range images.Afterwards, this processor 41 obtains arbitrary pixel in the attribute information, this each image subblock of each image subblock in these a plurality of image subblocks at brightness value of this high dynamic range images corresponding position etc., and wherein the attribute information of this each image subblock comprises the maximum L of the pixel intensity of each image subblock _Max, mean value L _Mean, variance L _StdDeng, according to arbitrary pixel in the attribute information of each image subblock in these a plurality of image subblocks, this each image subblock in the brightness value of this high dynamic range images corresponding position and the default linear mapping relations of image, adjust the brightness value of this arbitrary pixel in this each image subblock, brightness value according to each pixel in each image subblock after adjusting generates and sends low dynamic range echograms to this display 42.Adjust the pixel brightness value etc. of each image subblock by the cutting apart of 41 pairs of these high dynamic range images of this processor, zone mapping and process operation, can generate simply and effectively the low dynamic range echograms that comparatively approaches with this high dynamic range images display effect.

The variance L of above-mentioned each image subblock can also further be set by processor in addition, _StdBe 0.5, the maximum L of the pixel intensity of each sub-block _MaxAll according to mean value L _MeanObtain, specifically obtain formula with above-mentioned formula (2).

This high dynamic range images display unit can also comprise:

Memory is used for setting and stores multiple continuous and nonoverlapping range of luminance values reaching one to one brightness degree;

Processor, also be used for according to the range of luminance values under the mean value of the pixel intensity of each image subblock, obtain brightness degree corresponding to this each image subblock, according to the level spacing between brightness degree corresponding to adjacent image sub-block, adjust the maximum of the corresponding pixel intensity of image subblock that pixel brightness level is high in this adjacent image sub-block.

In embodiments of the present invention, after display shows this low dynamic range echograms, if when luminance difference is larger between adjacent image sub-block or the zone, because highly sensitive in colourity to brightness of human eye when then human eye is watched, easily produces sense of discomfort.Such as, in the actual displayed process, differ less adjacent image sub-block for average brightness, if adjust the L of latter two sub-block of pixel brightness value _MeanDifference is larger, then can produce false edge in procedure for displaying; And differ larger adjacent domain for average brightness, then in follow-up adjustment pixel brightness value process, should keep the edge that originally exists.The false problems such as edge appearance appears in the low dynamic range echograms after showing, can further optimize the display effect of this low dynamic range echograms.Also namely this memory by the embodiment of the invention, processor can head it offs, and concrete step sees that above-described embodiment two is described, does not repeat them here.

Embodiment four:

Fig. 5 shows the structure of the high dynamic range images display unit that fourth embodiment of the invention provides, and for convenience of explanation, only shows the part relevant with the embodiment of the invention.

This high dynamic range images display unit can be the software unit that runs in the terminal, also can be used as independently, suspension member is integrated in the terminal or runs in the application system of terminal, this high dynamic range images display unit comprises the first cutting unit 51, adjustment unit 52, image generation unit 53 and display unit 54, wherein:

This first cutting unit 51 is used for according to default image segmentation algorithm high dynamic range images being cut apart, and obtains and send a plurality of image subblocks to this adjustment unit 52;

This adjustment unit 52, be used for receiving these a plurality of image subblocks that this first cutting unit 51 sends, according to arbitrary pixel in the attribute information of each image subblock in these a plurality of image subblocks, this each image subblock in the brightness value of this high dynamic range images corresponding position and the default linear mapping relations of image, adjust the brightness value of this arbitrary pixel in this each image subblock, and the brightness value of each pixel is sent to this image generation unit 53 in each image subblock after will adjusting;

This image generation unit 53, brightness value for each each pixel of image subblock after this adjustment that receives these adjustment unit 52 transmissions, and according to the brightness value of each pixel in each image subblock after adjusting, generate and send low dynamic range echograms to this display unit 54; And

This display unit 54 is used for receiving this low dynamic range echograms that this image generation unit 53 sends, and shows this low dynamic range echograms.

In embodiments of the present invention, by this first cutting unit 51 according to the image partition method that pre-sets, original HDR image is formed a plurality of image subblocks after over-segmentation, this adjustment unit 52 can be added up three statistical attribute information obtaining each image subblock, also namely comprises the maximum L of the pixel intensity of each image subblock _Max, mean value L _Mean, variance L _StdAnd can be according to many local S curves of the corresponding generation of above-mentioned formula (1), the curve that also namely represents arbitrary pixel p mapping relations between the brightness value L (p) of this arbitrary pixel p in described each image subblock after the brightness value H of this high dynamic range images corresponding position (p) and the adjustment in each image subblock, because this curve is S-type, therefore can be described as the S curve.Then by each image subblock is carried out corresponding curve mapping, can realize the adjustment to the brightness value of all pixels in all image subblocks, further according to the brightness value after adjusting, can realize the low dynamic range echograms corresponding with this high dynamic range images, with this low dynamic range echograms after display unit 54 shows, reached make acquisition the LDR image more near the purpose of the actual acceptable effect of user.

In addition, be further to promote display effect, can also continue the attribute information adjusting the S parameter of curve or claim each image subblock in the embodiment of the invention, so that obtain better display effect after each sub-image shone upon.Such as, still obtain the attribute information of each image subblock, yet this moment L is set _MeanBe worth constant, L _StdValue is 0.5, L _MaxObtain according to above-mentioned formula (2), then pass through the mapping such as above-mentioned formula (1) after, the low dynamic range echograms of generation will have more excellent display effect,

In embodiments of the present invention, adopt the local mapping method of image region segmentation, can solve in the image local high/low brightness, the problem such as dizzy, regulate local parameter and can not affect other regional display effects, and the statistical property in local zone is combined with curve mapping method in the formula (1), only realize mapping process with the attribute information of each image subblock itself as parameter, and need not carry out the Adjustment operation of careful complexity, more be conducive to keep the characteristic of original image itself, the concrete better display effect of image that finally shows at display device.

Embodiment five:

Fig. 6 shows the structure of the high dynamic range images display unit that fifth embodiment of the invention provides, and for convenience of explanation, only shows the part relevant with the embodiment of the invention.

This high dynamic range images display unit comprises that compression unit 61, the second cutting unit 62, adjustment unit 63, image generate list 64 and display unit 65, and this adjustment unit 63, image generation single 64 and display unit 65 branches generate list 53 with adjustment unit 52 described in above-described embodiment four, image and display unit 54 is corresponding one by one, be not repeated herein, wherein:

This compression unit 61 is used for high dynamic range images is compressed, and obtains the low dynamic range echograms that display device can show, and this low dynamic range echograms is sent to this second cutting unit 62;

This second cutting unit 62 is used for by this low dynamic range echograms that receives these compression unit 61 transmissions, and according to default image segmentation algorithm, this low dynamic range echograms is cut apart, and obtains a plurality of image subblocks.

In embodiments of the present invention, pretreatment operation through the HDR image, also namely obtain the width of cloth LDR image demonstration of this HDR image by this compression unit 61, show the image segmentation algorithm that adopts based on graph theory by 62 pairs of these LDR images of this second cutting unit again, can realize the scene image of complexity is cut into the more consistent regional area of a plurality of internal features or a plurality of image subblock, and the segmentation result of the segmentation result of this width of cloth LDR image and original this HDR image is close.Because existing image partition method is all realized in the LDR image.The present invention at first uses the 0-1 algorithm to compress pretreatment operation one width of cloth HDR image and obtains the LDR image, again to this LDR image segmentation, can obtain better segmentation effect, can reduce amount of calculation and improve the speed of compression for the HDR image that contains mass data simultaneously, because processing speed is higher, the scene that the present invention processes in real time for needs has obvious advantage, as being applied in the scene such as wide dynamic video monitoring.

Behind the attribute information of each image subblock of the present invention after statistics is cut apart low dynamic range echograms, adjustment unit 63 is in conjunction with the attribute information of each image subblock, arbitrary pixel is in the brightness value of this high dynamic range images corresponding position and the default linear mapping relations of image in this each image subblock, realization is to the adjustment of the brightness value of this arbitrary pixel in this each image subblock, by the brightness value of this image generation unit 64 according to each pixel in each image subblock after adjusting, generate low dynamic range echograms, display unit 65 shows this low dynamic range echograms.

In embodiments of the present invention, after triggering adjustment unit 63, trigger before the image generation unit 64, this high dynamic range images display unit can also comprise:

Setting unit is used for arranging multiple continuous and nonoverlapping range of luminance values and reaches one to one brightness degree;

The grade acquiring unit is used for obtaining brightness degree corresponding to this each image subblock according to the range of luminance values under the mean value of the pixel intensity of each image subblock;

The brightness adjustment unit is used for according to the level spacing between brightness degree corresponding to adjacent image sub-block, adjusts the maximum of the corresponding pixel intensity of image subblock that pixel brightness level is high in this adjacent image sub-block.

In embodiments of the present invention, triggering after adjustment unit 63 realized brightness value adjustment to all pixels in all image subblocks, for fear of because when luminance difference is larger between adjacent image sub-block or the zone, the problem that causes the human eye discomfort occurs, can also be according to the difference of the average brightness between the adjacent image sub-block, further adjust the brightness maximum of image subblock, such as, the maximum of adjusting the corresponding pixel intensity of image subblock that pixel brightness level is high in this adjacent image sub-block for this adjacent image sub-block in the maximum etc. of the low corresponding pixel intensity of image subblock of pixel brightness level.Specifically realize the peaked optimization of image subblock brightness by setting unit, grade acquiring unit, brightness adjustment unit etc., its concrete performing step is seen described in above-described embodiment two, and then when realizing by high dynamic range images to the mapping of low dynamic range echograms by the linear mapping relations of default image, the display effect of the low dynamic range echograms that generates more meets the comfort of watching of human eye, and this low dynamic range echograms is more near original high dynamic range images, also so that display effect is better.

The embodiment of the invention is by carrying out Region Segmentation to high dynamic range images, the different image subblock that gets access to is used respectively mapping relations or claims mapping curve to process, be specially the attribute information according to each image subblock in a plurality of image subblocks after cutting apart, arbitrary pixel is in the brightness value of this high dynamic range images corresponding position and the default linear mapping relations of image in this each image subblock, adjust the brightness value of this arbitrary pixel in described each image subblock, finally be converted to the image in the low-dynamic range, the problem of local display will more directly be processed, and it is also fairly simple to adjust mapping curve corresponding to each image subblock, can not affect the display effect of other area image sub-blocks in the image, solved utilize prior art with the HDR image compression for the LDR image after, the problem that the LDR image display effect is not good, so that the low state range image that gets access to or claim the LDR image when display device shows, display effect is better.

The above only is preferred embodiment of the present invention, not in order to limiting the present invention, all any modifications of doing within the spirit and principles in the present invention, is equal to and replaces and improvement etc., all should be included within protection scope of the present invention.

Claims (11)

1. a high dynamic range images display packing is characterized in that, described method comprises:

According to default image segmentation algorithm, high dynamic range images is cut apart, obtain a plurality of image subblocks;

In the brightness value of described high dynamic range images corresponding position and the default linear mapping relations of image, adjust the brightness value of described arbitrary pixel in described each image subblock according to arbitrary pixel in the attribute information of each image subblock in described a plurality of image subblocks, described each image subblock;

According to the brightness value of each pixel in each image subblock after adjusting, generate and the demonstration low dynamic range echograms.

2. the method for claim 1 is characterized in that, described high dynamic range images is cut apart before, described method also comprises:

High dynamic range images is compressed, obtain the low dynamic range echograms that display device can show;

At this moment, describedly obtain a plurality of image subblocks and be specially:

According to default image segmentation algorithm, described low dynamic range echograms is cut apart, obtain a plurality of image subblocks.

3. the method for claim 1 is characterized in that, the attribute information of each image subblock comprises the maximum L of the pixel intensity of described each image subblock in described a plurality of image subblocks _Max, mean value L _Mean, variance L _Std

4. the method for claim 1 is characterized in that, the attribute information of each image subblock comprises the maximum L of the pixel intensity of described each image subblock in described a plurality of image subblocks _Max, mean value L _Mean, variance L _Std, wherein, described variance L _StdValue be 0.5, described maximum L _MaxObtain according to the following equation:

$L_{\max }={10}^5\×\left(\begin{array}{c}0.266\×{\mathrm{Lmean}}^6-1.102\×{\mathrm{Lmean}}^5+1.476\×{\mathrm{Lmean}}^4\\ -0.829\×{\mathrm{Lmean}}^3+0.205\×{\mathrm{Lmean}}^2-0.021\×\mathrm{Lmean}+0.001\end{array}\right).$

5. such as claim 3 or 4 described methods, it is characterized in that, the linear mapping relations of described default image are shown in the following formula:

$L\left(p\right)=\frac{L\max }{1+\exp (-\mathrm{Lstd}\×(H\left(p\right)-\mathrm{Lmean}))},$

Wherein, arbitrary pixel p is at the brightness value of described high dynamic range images corresponding position in described each image subblock of H (p) expression, and L (p) represents the brightness value of described arbitrary pixel p in described each image subblock after the adjustment.

6. the method for claim 1 is characterized in that, after the brightness value of the described arbitrary pixel of described adjustment in described each image subblock, before described generation and showing low dynamic range echograms, described method also comprises:

Multiple continuous and nonoverlapping range of luminance values is set reaches one to one brightness degree;

According to the range of luminance values under the mean value of the pixel intensity of each image subblock, obtain brightness degree corresponding to described each image subblock;

According to the level spacing between brightness degree corresponding to adjacent image sub-block, adjust the corresponding pixel of image subblock that pixel brightness level is high in the described adjacent image sub-block bright.

7. a high dynamic range images display unit is characterized in that, comprises the first cutting unit, adjustment unit, image generation unit and display unit, wherein:

Described the first cutting unit is used for according to default image segmentation algorithm high dynamic range images being cut apart, and obtains and send a plurality of image subblocks to described adjustment unit;

Described adjustment unit, be used for receiving described a plurality of image subblocks that described the first cutting unit sends, reach the linear mapping relations of default image according to the brightness value of arbitrary pixel in the attribute information of each image subblock in described a plurality of image subblocks, described each image subblock in described high dynamic range images corresponding position, adjust the brightness value of described arbitrary pixel in described each image subblock, and the brightness value of each pixel is sent to described image generation unit in each image subblock after will adjusting;

Described image generation unit, brightness value for each each pixel of image subblock after the described adjustment that receives described adjustment unit transmission, and according to the brightness value of each pixel in each image subblock after adjusting, generate and send low dynamic range echograms to described display unit; And

Described display unit is used for receiving the described low dynamic range echograms that described image generation unit sends, and shows described low dynamic range echograms.

8. device as claimed in claim 7 is characterized in that, described device also comprises compression unit and the second cutting unit:

Described compression unit is used for high dynamic range images is compressed, and obtains and send low dynamic range echograms that display device can show to described the second cutting unit; And

Described the second cutting unit is used for receiving the described low dynamic range echograms that described compression unit sends, and according to default image segmentation algorithm, described low dynamic range echograms is cut apart, and obtains a plurality of image subblocks.

9. device as claimed in claim 7 is characterized in that, the attribute information of each image subblock comprises the maximum L of the pixel intensity of described each image subblock in described a plurality of image subblocks _Max, mean value L _Mean, variance L _Std

10. device as claimed in claim 7 is characterized in that, the attribute information of each image subblock comprises the maximum L of the pixel intensity of described each image subblock in described a plurality of image subblocks _Max, mean value L _Mean, variance L _Std, wherein, described variance L _StdValue be 0.5, described maximum L _MaxObtain according to the following equation:

$L_{\max }={10}^5\×\left(\begin{array}{c}0.266\×{\mathrm{Lmean}}^6-1.102\×{\mathrm{Lmean}}^5+1.476\×{\mathrm{Lmean}}^4\\ -0.829\×{\mathrm{Lmean}}^3+0.205\×{\mathrm{Lmean}}^2-0.021\×\mathrm{Lmean}+0.001\end{array}\right).$

11. such as claim 9 or 10 described devices, it is characterized in that, the linear mapping relations of described default image are shown in the following formula:

$L\left(p\right)=\frac{L\max }{1+\exp (-\mathrm{Lstd}\×(H\left(p\right)-\mathrm{Lmean}))},$

Wherein, arbitrary pixel p is at the brightness value of described high dynamic range images corresponding position in described each image subblock of H (p) expression, and L (p) represents the brightness value of described arbitrary pixel p in described each image subblock after the adjustment.

Images