TWI295895B - Adaptive color adjustment method and device for digital images in a display apparatus - Google Patents

Description

1295895 1863] *twf; doc7r IX, invention description: [Technical field of the invention] The invention is related to the display of 11 digital image adaptive color ^ and f ^ ° This invention has a special _ an adjustable image The brightness of the person is the adaptive color of the shadow square n device. [Prior Art] Recently, the display has placed considerable emphasis on the visual quality of the output image. The hardware interface also provides 0SD (0nscreenDisplay) for the user to adjust the output image quality.

Quality, so the display does not show the operation - OSD refers to the function of adjusting the volume or brightness, the face will be displayed in a plaid. Today's (10) integrated functions are equivalent to ^, in addition to basic brightness, contrast, shift adjustment and other functions, even including barrel distortion, trapezoidal distortion, parallelogram distortion, acupuncture phenomenon, etc. can be adjusted 〇SD Set the value to improve. The advantage of using OSD is that the knob is greatly reduced, which reduces the chance of poor contact, and can also save the adjusted settings. However, the QSD setting is set to -gj, and the display is always _image The display materials are optimized. .......... There are also several conventional image adjustment techniques, which are listed below. US Patent No. US6778183 B1 uses the probability distribution of the accumulated image brightness signal to determine the image. The basis of the overall brightness attribute. According to this method, the attributes of the image are classified into dark images, dark tone images or dark images (dark, mid tone or bright image), or attributes of any two combinations. The property can adjust the image contrast by using three different input and output correspondence tables 1295895 1863itwf.doc/r (In-Out LookUp Table). At the same time, the basis of the dynamic range of the dynamic range is defined by the agricultural dark point and the brightest point in the image pixel. The focus of this patent is to provide two user-changeable LUTs to enhance contrast and dynamic range. The feature is the use of look-up tables, which is fast.

U.S. Patent No. 6,771,837 B1 discloses a method of improving the overall contrast of an image. By using the frequency distribution of all the image information distributions in the spatial domain to the frequency domain, the Fourier transform method is used to distinguish the high frequency and the low frequency, and then the image content is captured. Its characteristics are significant, but it is quite complicated due to the complexity of the Fourier transform method when the hardware is implemented. 〇秀国导利号US6618045 Β1 exposes - a display cry # set, the user's ambient light sensor; - group (10) can be mentioned = user slaves like the screen pillow brightness, shame; sigh special treatment ^ 耆Week ^ ̄ ̄ ̄ ̄ ̄ ̄ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ Simple health mechanism material = like == calculation burden, also highlights the difficulty of hardware circuit design ^ ^ private not only causes the end of the operation: the image adaptation method of the management of the good, can be improved, the appearance of the invention The purpose of the present invention is to provide an image-adjusting color adjustment 1295895 1863Itwf.doc?r mechanism's original image brightness information of the luminance component of the statistical image signal, and recalculating the probability distribution of the human eye to the bright household. See the automatic balance of the child's degree after the other - a 3 mechanism of the invention, according to the _ count ^ ^ = difficult mechanism, according to _ count ^ | = mention / ^ 4 image touch color adjustment number, select 浑 、, this is after I The branch/knife map calculation adjusts the image contrast to construct the image contrast enhancement curve of the f-bank.另一 Another purpose of X month is to provide an image-adaptive color ▲ ▲ 机Niu low field should change the mechanism ^ color adjustment u 4 chromaticity image signal at least has a brightness component and ^ color component to read the unit to analyze and recalculate the brightness of the chrominance image signal = degree of probability distribution The graph 'to obtain a - weight luminance probability distribution map Γ - a dynamic range luminance correspondence table generating unit receives the round of the analysis unit. The weight distribution map of the weight of the sea is generated, and the dynamic range brightness corresponding to the __ is generated. The pair of enhanced brightness correspondence table generating unit receives the weighting probability distribution map of the weighting unit of the analyzing unit, and generates a contrast enhanced brightness correspondence table. For adjusting the image contrast of one of the brightness components; a brightness adjustment unit, skirt 1295895 1863itwf.doc/r according to the dynamic range brightness correspondence table and the contrast enhanced brightness correspondence table, to adjust the brightness component of the brightness chrominance image signal A chroma adjustment unit generates a chroma compensation factor for compensating for the chroma component of the luminance chroma image signal according to the brightness relationship before and after the adjustment. Thereby, the image signal obtained by the luminance component and the chrominance component is obtained. A further embodiment of the present invention provides a digital image adaptive color adjustment device for a display, comprising: a signal conversion unit that receives a first φ image signal for conversion into a second image signal, the first image signal is not The component has a luminance and chrominance separation component, and the second image signal has a component of luminance and chrominance separation; an analysis unit analyzes and weights and calculates a probability distribution diagram of the shell component of the second image k In order to obtain a weight luminance probability distribution map; a dynamic range brightness correspondence table generating unit receives the weight luminance probability distribution output by the sub-unit, and generates a dynamic range party degree correspondence table for adjusting the brightness component. a dynamic range; a contrast redundancy correspondence table generating unit receives the weight luminance probability distribution outputted by the analyzing unit, generates a contrast enhancement brightness correspondence table, and adjusts the brightness component-image contrast; The machine unit takes a plurality of dynamic range brightness correspondence tables generated by the dynamic L range exemption correspondence table generating unit, and smoothes the 1 after sliding, the dynamic brightness correspondence table, and taking the contrast increase (four) degree correspondence table generating unit, the plurality of pairs of pairs of enhanced brightness, and smoothing, to generate - smoothing and contrasting enhanced party correspondence table; - brightness adjustment The unit adjusts the brightness secret surface of the second image signal according to the smoothed dynamic range brightness correspondence table and the smoothed contrast enhanced brightness correspondence table; a color ^ 1295895 18631twf.doc/r compensation The unit generates a lottery number according to the brightness relationship, and is used to compensate the chrominance component of the second image signal; and the converting unit receives the chrominance component of the second image signal and the tone component In order to convert into an output image signal, the round-out scene image 4 < has a luminance, chromaticity separation component. . A further embodiment of the present invention provides an image adaptive color method, comprising: (Α) converting a continuous input-transmission shadow_number into a luminance chrominance separation signal, the luminance chrominance separation signal comprising a bright sound (B) statistically calculate the probability distribution of the luminance component, and perform adaptive allocation to obtain a -weight luminance probability distribution map; (c) dynamic range (DR) extension of the weighting rate distribution map, To obtain a dynamic bear correspondence table (LUT-DR); (D) the luminance luminance distribution enhancement (CT) to obtain a contrast enhanced brightness correspondence table (LUT_e curve smoothing the contrast enhanced brightness correspondence table; (F) adding a plurality of dynamic range luminance correspondence tables corresponding to the 'Ancient () component of the complex text to a dynamic range brightness correspondence table of one subsequent luminance component, and ^ 肷 a plurality of luminance components Corresponding plural smoothed contrast enhanced highlights are added to the i table to form a _ pair = two _ = degree correspondence table applied to the subsequent brightness component; (G) the dynamic range brightness corresponding to the average is compared ^ = contrast Enhance the brightness correspondence table for the subsequent brightness The composition is adjusted ^ (H) according to the degree of difference between the brightness components before and after the adjustment, ^ 2, the compensation factor; (I) according to the color compensation factor generated, the shadow compensation; and (7) Performing inverse conversion on the brightness adjustment and the chrominance adjusted split signal to produce a continuous output image signal 1295895 18631twf.doc/r To make the above and other features and advantages of the present invention more apparent, DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS The embodiments of the present invention are described in detail below with reference to the accompanying drawings. [Embodiment] Embodiments of the present invention disclose an analysis of input image content, calculating parameters for adjusting contrast and dynamic range, and generating brightness corresponding FIG. 1 is a functional block diagram showing chroma compensation according to an embodiment of the present invention. The function of the adaptive color adjustment device 100 includes at least: performing signal conversion to obtain brightness. Chroma separation signal; calculate the parameters of the dynamic range (Dynamic Range, referred to as DR) and automatically generate the brightness correspondence table (LUT of dynam Ic range stretch (LUT-DR abbreviation); calculate the parameters of the contrast (Contrast, CT abbreviation) and automatically generate (or access to the built-in database) contrast enhanced party correspondence table (LUT 〇f contrast Enhancement, to LUT-CT abbreviation); to increase the smoothness and adjust the overall brightness of the image, the LUT_CT curve can be automatically adjusted slightly. If the input image δίΐ is the silent image *5 tiger, according to the image party component The degree of difference between the automatically generated LUTJDR and the LUT-CT adjusted brightness component of the automatic fine adjustment is calculated, the chroma compensation factor is calculated, and the luminance and chrominance components of the image signal are adjusted to output a digital image. If the input image signal is a continuous moving sad image signal, for the purpose of output image comfort, a plurality of automatically generated LUT-DR and the fine-tuned LUT-CT are processed for smoothness, according to the brightness component before and after the adjustment. To determine the degree of difference, calculate the chroma 10 1295895 1863 rtwf.doc/r compensation factor, and adjust the brightness and chrominance components of the image signal to output a digital image. The adaptive color adjustment device 100 shown in FIG. 1 includes: a signal conversion unit 110, an analysis unit 115' dynamic range brightness correspondence table generation unit 120, a contrast enhancement brightness correspondence table generation unit 130, and a smoothing adjustment unit 140. Brightness adjustment unit 150, chroma compensation unit 170, and inverse signal conversion unit 160 °

The signal conversion unit 110 receives an input video signal (RGB) and converts it into a luminance chrominance separation signal such as YCbCr, YUV, HIS, YPbPr and the like. In the present embodiment, the input image signal (RGB) must be a digital image, which can be a static display image or a dynamic continuous playback movie. Figure 1 shows only YCbCr as an example. In a general 8-bit processing system, the value of RGB is a positive integer between 0 and 255, and the range of YCbCr is 0 to 255' Cb, Cr: -127 to 127 y YCbCr are integer values. The signal converted by the signal conversion unit 110 is used to analyze the image content and calculate the adjusted tea number. If the wheel image is already a signal of luminance and chrominance separation, the signal conversion unit 110 may not be required. The analysis list S 115 is used to calculate the histogram of the brightness γ and to simulate the automatic brightness balance mechanism of the human eye. The analysis unit ιΐ5 is divided into the brightness Y, and the statistics of all the images in the luminance side are distributed. In: m is used as the means of brightness probability distribution, and the brightness is distributed according to the total number/value range (ie, the brightness range of the 8-bit system is 0~255; 1_down, 0~1023). The resulting luminance probability distribution is shown by the solid curve in Figure 2. 1295895 18631fwf.doc/r Furthermore, in order to simulate the automatic balance of the brightness of the human eye, the analysis unit 115 more appropriately assigns the luminance probability profile. = If the probability ratio is higher than the number of occurrences (Μ_), the ratio is adjusted; if the number is lower than MaxH/2, it remains unchanged. As shown in Fig. 2, hl >MaxH/2 adjusts Μ to hi, ' hl, = (MaxH/2+hl)*0 5 . If the number of times is lower than MaxH/2, the original number of occurrences, such as hO< MaxH/2, remains unchanged. According to this adjustment method, a luminance probability distribution map (Adapjiv'Hist〇gram, referred to as ADH) which is different from the original image can be obtained, as shown by the dotted line in FIG. The ADH generated by the analyzing unit 115 is input to the dynamic range brightness correspondence table to generate a single it 12G and contrast enhanced brightness correspondence table generating unit 130 for generating the moving evil range party degree correspondence table (luT_dr) and the contrast enhanced brightness correspondence table. (LUT_CT;). A dynamic range luminance correspondence table generating unit 120 performs a dynamic range (DR) extension processing on the ADH generated by the analyzing unit 115. First, the 'dynamic range brightness correspondence table generating unit 12' calculates the original black_(by DRb_) and the self-order point (referred to as DRw for short) based on adh. The DRb can be the digital value of the darkest point of the image towel, or can be accumulated on the ADH curve from the number of recordings to the high-touch direction to a certain value (such as the digit value of (U). The DRw can be in the image. The digit value of the brightest point can also be accumulated on the ADH (four) from the number 255 to the lower digit value to the number of occurrences of a certain value (such as 〇1). For example, in the wrist system, the ADH distribution map of an image, The total occurrence probability of the accumulated digit value 〇~1〇 can reach the cumulative probability to the division αι, then which can be determined = 12 1295895 18631twf.doc/r 10; the total occurrence probability of the accumulated digit value 230~255 can reach the cumulative probability greater than or equal to 0 ·1, then DRw = 230. The maximum degree of DR should be distributed in the whole digit range of the system, namely: in the 8-bit system, DRb = 0, DRw 2 255; in the i〇bit system, DRb = 0 ' DRw = 1023. After calculating the DRb and DRw of the original image by the above method, the digital value DRb can be arbitrarily defined to be smaller than DRb, and (DRb, < DRb) is a new black level point 'arbitrarily defining a digital value DRw' larger than DRw ( DRw'> DRw) is defined as a new white point. Calculate DRb, DRw and define DRb', DRw, The purpose is to calibrate the two turning points of LUT-DR. The turning point is; the establishment of the coordinate points on LUT-DR is (DRb, DRb,) and (DRw, DRw,), where DRb and DRw are the black level of the original image. Point and white point; DRb, and DRw are new black point and new white point. The schematic diagram is shown in Figure 3. The dotted line is the LUT curve without DR extension adjustment, and the solid line is adjusted by DR extension. The LUT curve 0 can automatically establish the output value (〇utput) corresponding to the input value (Input) on the LUTJDR curve according to the two end points and the two turning points. The output value can be generated by linear interpolation or nonlinear interpolation. The method of calculating the degree of the knife map is to complete the DR extension operation according to the table. The contrast enhanced brightness correspondence table generating unit 130 performs contrast enhancement processing on the ADH. According to the ADH, the image dark color and the bright color pixel are further calculated into 8 bits. Under the system, the scope of the dark component is defined on the probability distribution map on the digit value 0 to 40; the range defining the bright component is distributed on the probability map on the digit value 215 to 255. Under the l〇bit system, the dark color Minute

13 1295895 18631twf.doc/r is distributed from 0 to 160; the party color is distributed from 863 to 1023. - In the present embodiment, the cumulative probability value (indicated by F1 for this cumulative probability value) and the cumulative probability value within the bright color range (indicated by F 2 for this cumulative probability value) are calculated. Different F1 and F2 values have dark and bright color components, and the LUT-CTM with different curved shapes is required to enhance the visual contrast. In the present example, the LUT-CT is composed of the built-in databases LUT_CTF1 and LUT-CTF2, and its coverage is shown in Fig. 4. According to the F1 value φ, the half-sigmoid curve (LUT-CTF1) of the dark component must be enhanced; the half-sigmoid curve (LUT-CTF2) of the bright component must be selected according to the value of F2. In the 8-bit system, LUT-CTF1 is a bone correspondence table with four digit values 〇 to 79 (a, b, c and φ in Fig. 4; LUT-CTF2 is a four-digit value 80 to 255 对应 degree correspondence table (Figure 5 • e, ^, g and . Under lObit system 'LUT-CTF1 is the four-digit value 〇 to 317 brightness correspondence table, LUT-CTF2 is four digit values 321 to 1〇23 The brightness correspondence table. The selection conditions of the four LUT-CTF1 can be determined according to the calculated F1 and k* (F1 calculation range/full brightness digit value); the selection conditions of the four lut_ctf2 can be calculated according to the calculation. !^ * (]? 2 calculation range / full redundancy digit value) to determine. k疋 select the adjustment factor of the semi-curve curve, the 匕 value is between 0~1·2. Take the 8-bit system as an example, F1 calculation range 4〇一〇= 4〇; F2 calculation range is 255 -215 = 40; full brightness digit value is 255 〇 = 255. Therefore, only according to the F1 and F2 values, the selection table can be selected to enhance Comparative LUT-CTF1 and LUT-CTF2 curves. Half s shape 14

(I 1295895 18631twf.doc/r Curve selection conditions are listed according to the different & values listed below:

Semi-sigmoid curve selection condition

LUT CTF1 (4).16)>?1>^^)

Kl = 1.2 k2 = 0.8 k3 = 0·5 and LUT-CTF2 curve. However, if the selected <k value LUT-CTFl LUT CTF1 and LUT_CTF2 curves are directly combined into LUT_CT, in some cases, the slope of the curve at the junction point of LUT-CTF1 and LUT_CTF2 may be excessively changed. 'Cause some images to appear uncoordinated after the LUT-CT brightness is checked. In order to reduce the occurrence of uncoordinated pupils while achieving image brightness optimization, the present embodiment increases the smoothness of the LUT-CT curve in the following manner. For example, as shown in FIG. 5, the digit values of the LUT-CTF1 and LUT-CTF2 junction points (referred to as midgry abbreviation) are respectively subtracted from the first fine-tuning coefficient gl and the second fine-tuning coefficient g2, and the midgry-gl, midgry and Midgry + g2 Three-pronged punctuation for fine-tuning of brightness enhancement. 15 I2958953itwfd〇c/r T—Under the system, the boundary point between LUT-CTFl and LUT-CTF2 is 8增加 to increase the overall smoothness of LUT-CT. After LUT-CTF1 soldier and CTF2 construct LUT-CT, take g Bu 10, g2 2, 20, ::70 80 1 〇〇 two punctuation points for brightness correction. The purpose of the correction is to make the curve between 70 and 100 smoother.

The L square can be used to take the 7-to-1D LUT-CTR value for the sub-linear interpolation of 7 G and (10) two punctuation points; or to reconstruct 70 to 1 by linear interpolation with the temple 3 f punctuation. . . # LUT_CT Content value, as shown in Figure 5. The LUT-CT obtained in this way has the function of strong image contrast and image brightness. In Fig. 5, the original selection, LUT_CTF1 and LUT_CTF2 are shown. As for the punctuation point %~1〇〇, then fine-tune the area. In summary, for the purpose of enhancing image contrast and image brightness optimization, the contrast table of the enhanced brightness correspondence table generating unit 13 and FIG. 5 is compared. In the digital image of continuous playback, this embodiment can generate different LUTJDR and LUT CT for the image content of the individual frame. If the difference in LUT between the binocular frames is too large, it is easy to visually create an uncomfortable feeling. At the same time, when the hardware is implemented, the loss time of the hardware operation enhancement parameter must be considered. The smoothness adjustment unit 14〇 can be used for color smoothing (C.S.), and its schematic diagram is as shown in FIG. 6. ‘It is assumed that the number of hardware transports (ie, the correspondence table) must take up two frames. In this embodiment, the smoothing mechanism of C.S. is to take the sum of the values of the curve contents of the LUTs that have been shipped to take the 16 1295895 18631twf.doc/r LUT of the view. In more detail, the lUTJDR applied to the Nth to N+2 frames is the intra-valley phase of the LUT-DR obtained from the N-6th frame and the L_DR obtained from the n_3th frame. Plus average. Similarly, the LUT-CT applied to the Nth to N+2 frames is the result of the LUT-CT obtained from the N_6th frame and the service result of the first frame. 2 averaged. In Fig. 6, LUT_N_6 represents LUT-DR and LUT-CT which are known from the N_6th frame, and other meanings can be analogized. According to the LUT DR, LUT_CT averaged by the smoothness adjusting unit 140, the redundancy adjusting unit 15 adjusts the luminance signal γ to obtain a new luminance signal Y'. The shirt compensation sheet το 17 0 generates a chroma compensation factor f according to the relationship between the original luminance signal γ and the new luminance signal Y'. The chroma compensation factor f , (YVY) may be a linear relationship or a multiple linear relationship. Chroma compensation early π Π 0 and then compensate the luminance chromaticity according to the chroma compensation factor W generated. The chrominance signal (Cb, Cr) of the spring separation signal, and the chrominance signal (Cb, Cr') after compensation. The relationship between the image-the chromaticity signal (10) (7) and the chromaticity of the image 2 (Cb, Cr,) can be as follows:

Cb9 = f*Cb ;

Cr, = f * Cr ; After taking the inverse signal conversion unit 16 彳亍 according to the new luminance signal γ, and the complement: post chrominance signal (Cb ', cy), 彳亍 inverse conversion to obtain the output image = (R 'G'B'). Output image signal, 18,) can improve image contrast, sway and image smoothness, while maintaining image color consistency. 17 1295895 18631twf.doc/r Note that if the input image is a still image (such as a digital photo), etc., the embodiment of the present invention (such as item 1) may not include the smoothness adjusting unit 14A. That is, the correspondence tables LUT_fR and LUT_CT generated by the correspondence table generating units 120 and 13〇 directly turn to the luminance unit 15〇. Weekly redundancy § § Y, but still requires the chroma compensation unit 17 〇. Another embodiment of the present invention provides an image adaptive color adjustment method. The rounded image signal is an RGB digital signal. It is assumed that the hardware operation color φ has to take up two frames. The method of this embodiment then includes the following steps. A.) Convert all the digital signals RGB of the Nth frame in the image to the signal YCbCr separated by luminance and chrominance. B·) Statistics of the probability of occurrence of luminance γ in the entire image of the Nth frame. C·) Adaptive weight assignment (adh) to the original image probability distribution map. D·) Perform dynamic range (DR) extension·· Take the minimum digit value appearing in ADh as DRb (for example, set to 30), and set the maximum digit value to be 'DRw (for example, set to 240), arbitrarily set DRb, (for example, set to 1〇) and DRw (such as 'set to 250), where DRw, <DRb; DRw, > DRw. According to the coordinate point (〇, 〇) (3〇, 1〇) (240, 250) (255, 255), the LUT-DR ’ is calculated from four points, and the output digit value is calculated by linear interpolation or nonlinear interpolation. E·) Contrast enhancement: Calculate the F1 value of the dark component (digit value 〇~4〇) under ADH (such as '〇·〇5), and the F2 value of the bright component (digit value 215~255) (such as 'for 〇 · 3). According to the F1 value, the semi-f-curve curve of LUT-CTF1 model 18 1295895 18631twf.doc/r is used in the database; according to the F2 value, an appropriate half-sigmoid curve is selected from the database in the range of -, so that the LUT can be constructed. —CT. In order to enhance the power of the mouth-CT curve smoothness and image brightness optimization, the two punctuation points of the smooth fine-tuning area (7〇~1〇〇, its selection method ^ above) on the LUT-CT are in a linear manner ( Or take the three-point punctuation in a quadratic linear manner) 4 new calculation of the digit value 70 to 1G0 _ moment _ capacitance to complete the LUT-CT fine-tuning procedure. #F·)Improve the image smoothness: add the average content of the LUT_DR obtained from the N-6th frame to the LUT-DR of the second to the third frame to become the Nth to N+2 views. The LUTj[)r that must be enhanced by the frame; the LUT-CT obtained from the n-6th frame and the LUT-CT obtained from the frame are averaged to become the Nth to N+2 views. The frame must be enhanced by LUT-CT. G·) Query the brightness signal of the original image (assuming the digital value of a certain luminance chrominance signal in the image (Y, Cb, Cr) = (1〇〇, 3〇, _30)), and query the LUT-DR according to the table method. The LUT-CT can obtain a new luminance signal (assuming the digital value of the new luminance chrominance signal (Y,, Cb, Cr) = (11 〇, 3 〇, _30)). H·) According to the relationship between the original luminance signal (γ) and the new luminance signal (γ,), the chroma compensation factor is defined (assuming a linear relationship between f and γ, /γ, that is, then 110/100 =1.1), the chrominance component (Cb, Cr) of the original luminance chrominance signal is appropriately scaled up or down according to the f value (ie (Cb', Cr') = (33, _33)). I·) Convert the adjusted luminance signal (Y,) and the chrominance signal (Cb', Cr') back to the RGB signal, and output the image to complete the image processing program. 19 1295895

. Since the calculation and the judgment can be obtained, the object of the adaptive color adjustment is not more suitable. In the embodiment of the present invention, the static human eye can be more comfortable and natural, and the color tone consistency can be maintained before and after the image adjustment. By the same image as the content of the LUT, . The present invention has been disclosed in the above preferred embodiments, but it is not intended to be invented, and any skilled person can make some modifications and (4) without departing from the spirit and scope of the present invention. The scope of protection depends on the application of the gamma _ defined by the scales. BRIEF DESCRIPTION OF THE DRAWINGS Fig. 1 is a functional block diagram of an adaptive color adjustment device 1 according to an embodiment of the present invention. Figure 2 is the image brightness distribution map and weight distribution image brightness distribution map; thousand Figure 3 is the image dynamic range adjustment (LuTJDR); Figure 4 is the image contrast adjustment (LUt_CT) diagram; Figure 5 is the enhanced image contrast adjustment curve Schematic diagram of smoothness; Figure 6 is a schematic diagram of improving image smoothness. [Description of main component symbols] 1〇〇: Adaptive color adjustment device 1295895 18631twf.doc/r 110: Signal conversion unit 115: Analysis unit 120: Dynamic range luminance correspondence table generation unit 130: Comparison enhanced luminance correspondence table generation unit 140: Smoothness adjustment unit 150: brightness adjustment unit 170: chroma compensation unit 160: inverse signal conversion unit

Claims (1)

1295895 18631fwf.doc/r X. Patent application scope: 1. A digital image adaptive color adjustment device for displaying a person, comprising: an analysis unit 接收 receiving a luminance chrominance image signal, the luminance chrominance image signal having at least - The luminance component and the chrominance component, the analysis unit analyzes and reproduces the probability distribution map of the luminance component of the party chroma image signal to obtain a weight luminance probability distribution map; - a sorrow fen party degree correspondence table generating unit, receiving the "luminance solution distribution map outputted by the analysis unit, and performing dynamic range extension table to generate a dynamic range brightness correspondence table; - contrast enhancement exemption correspondence table generation Unit, connected to (4) analyze the single output of the "brightness scale distribution κ, contrast enhancement, raw-contrast enhanced brightness on the silk, used to compare the brightness components; m,, the whole single 70 ' according to the axial range brightness Forming and enhancing the brightness correspondence table to wrap the portion of the luminance chrominance image signal; and the war-color compensation unit to compensate the chrominance of the luminance chrominance image signal according to the brightness component between the front and the back Ingredients. The 3rd number is converted into a chromaticity chromaticity and the chromaticity of the chromaticity of the chromaticity. 2. If you apply for a spectator, you can adjust the image adaptive adjustment device, and you can input the image. Letter 22 1295895 18631twf.doc/r, 敕 1.72 of the digital image adaptive color adjustment garment described in item 3, further comprising an inverse signal conversion unit for receiving the adjusted brightness component output by the brightness adjustment And the signal conversion unit 该 the other component of the luminance chrominance image signal: one of the copies of the tU image signal. The shirt does not have a degree of image-adapted color as described in item 1 of the second dimension. The image of the luminance chrominance image includes: Yc YUV signal, HSI signal or YPbPr signal. The digital image adaptive color described in item 1 of the second application 模拟 心 模拟 simulates the adaptation of the human eye to the image of the f. 7 such as the weight perception probability distribution of the perception perception. The digital image adjusting device according to item 1 of the patent scope, wherein the dynamic range brightness corresponds to the table production, the - darkest point digital value and the - most bright two;; ^ a younger - black level point value and - the first - white point point value, to use one of the second black point point value and the - second white point point value to force $ > wide 7 - black point point value is less than the first - black level point number two : and the second white point point value is larger _ first - white point point value, = - black point point value and the second white point point position range brightness table. The digital image adaptive color range brightness correspondence table generating unit described in item 1 of the Lejia ϊli Li Fan Park will accumulate the probability value of the insertion weight exemption probability map up to the first number (? 23 1295 Lin - the responsibility digit The value defines a -th order - black point 'point value and - a first white point point number value for defining the extension - dynamic range - the second black level point value and the - second white point point value, (4) the second black level point digit value is smaller than the first black level point digit value 'and the second white level point bit value is greater than the first white level point digit value, the second black level point digit value and the second white point value The bit value of the order point is used to generate the dynamic range brightness correspondence table. 9. The digital image adaptive color adjustment device of the seventh embodiment of the claim 4 (4) is a linear interpolation method. Or the non-linear interpolation method is calculated. 10. The digital image adaptive color adjustment device described in the article Wei Wei, Item No., the towel ratio is enhanced, and the county generating unit calculates the weight exemption probability map. - Dark component probability cumulative value (four) and a bright color The cumulative rate of the extension rate (F2); according to the cumulative value of the dark component (F1) and the cumulative value of the bright component (F2), the contrast enhancement highlights the two groups covering the range of different digit values. ^ Correspondence table (LUT-CTF丨) and the second part contrast enhanced brightness pair ^ table (^υτ~〇τ2) constructed, the first part contrast enhanced brightness correspondence table and the second part contrast enhanced brightness correspondence table A digital image adaptive color adjustment device according to claim 10, wherein the contrast enhancement enhanced brightness correspondence table generating unit is added to increase the contrast enhancement brightness correspondence table. According to the first part of the contrast enhanced brightness correspondence table (LUT-CTF1) and the second part of the contrast strongness correspondence table (LUT_CTF2), a fine adjustment area ^ curve line smoothness fine adjustment is obtained. 24 1295895 18631twf.doc/r Color tone 3 set ^= Range 2 items of digital image adaptive color use - sub-linear interpolation or multiple linear interpolation square fine] Color tone = 4 special fine digital image fiber composition Difference between jobs The hue - a consistency ten iso - saturation compensation factor, to maintain Movies ^ 14.···································································· Multiple linear relationship ^ • · Digital image adaptive color adjustment device, including: - Signal conversion unit 'receives' the first image signal to be converted into two shadows, the signal, the first image signal does not have - a brightness component, the second image k has at least one luminance component and one chromin component; an analysis unit analyzes and weights a probability distribution map of the luminance component of the second image signal to obtain a weight luminance probability distribution map a dynamic range brightness correspondence table generating unit receives the weighted luminance probability distribution image output by the analyzing unit, and performs dynamic range extension to generate a dynamic range brightness correspondence table; a contrast enhanced brightness correspondence table generating unit receives The weighting probability distribution map of the weight of the analysis unit is outputted, and the contrast enhancement is performed to generate a contrast enhanced brightness correspondence table. Adjusting the image contrast of one of the brightness components; 25 1295895 18631twf.doc/r A smoothness adjustment unit, taking a plurality of dynamic range brightness correspondence tables generated by the dynamic range brightness correspondence table generating unit, and smoothing to generate a smoothed a dynamic contrast range brightness correspondence table, and a plurality of contrast enhanced brightness correspondence tables generated by the contrast enhanced brightness correspondence table are generated, and smoothed to generate a smoothed contrast enhanced brightness correspondence table; a brightness adjustment unit, according to the smoothed dynamic range brightness correspondence table and the smoothed contrast enhanced brightness correspondence table, to adjust the brightness component of the second image signal to generate an adjusted brightness component; a chroma complement unit, Compensating for the chrominance component of the second image signal according to the relationship between the brightness components before and after the adjustment; and an inverse signal conversion unit receiving the compensated chrominance component and the _then luminance component of the second image signal, In order to convert to an image signal, the output image signal does not have a luminance component. 16. For example, the application for the Wei Wei 15th excerpt image adaptive color adjustment device, wherein the second image signal comprises: YCbCr signal, γυν signal, HSI signal or YPbpr signal. For example, the touch image adaptive color adjustment device described in the application fiber M 15 is wherein the first image signal is a motion image. The invention relates to a digital image color adjustment device according to claim 15 , wherein the analysis unit simulates a human eye to visually perceive the visual image of the image, and the probability distribution map of the brightness component is performed. Day: Appropriate weight distribution to obtain the weighting degree map corresponding to visual perception. And the distribution of the Ge屮 19. The digital image of the object of the application of the scope of the patent range 26 1295895 18631t*wf.doc / r H of the towel axis range brightness correspondence table produces a single it the weight of the party, One of the darkest point digit values and one of the brightest point digit values is defined as a -th-black-order point digit value and a white-order point digit value, arbitrarily defined as one of the second-black point points of the extension-kinetic range The digit value and a second white point point value /, the medium U 苐 黑 - black point point 娄 field value is smaller than the first black level point value, and the second white point point value is greater than the first white level The point number value, the second black point point value and the second white point point value are used for the state range brightness correspondence table. ^ 20· The digital image adaptive color adjustment device according to item 5 of the patent application scope, the axial state brightness to the silk generating unit will accumulate the probability value of the weight brightness probability distribution map - the first value and the - The two-digit party value value defines a bit value and a first white-order point value value for defining an extension-dynamic norm value r a second white-order point number value, wherein the second black is at 5 haidi 1 The step digit value, and the second white point point digit value is greater than the first white point point digit value, the second black order point digit value and the second white order point digit value secret axis _ 目亮麟县. 21. The application of the dynamic range brightness table according to the numerical interpolation method or the nonlinear interpolation method is as follows: 彳f(4)15 items Kawasaki Touch Image Adaptive Color Weight Heavy Brightness Probability Map - Dark Component Probability Accumulated Value (4) and ^ Color Component Probability Accumulated Value (F2), 根据 based on the cumulative value of the dark component probability) 27 1295895 18631tWf.doc/r and the bright component probability accumulation value (F2), the contrast enhanced brightness correspondence table is composed of two sets of the first part of the range covering different range values. ^^,,,,, (LUT-CTF1 ) Contrast with the second part of the enhanced brightness ^ = (LUT-CTF2), the first part of the contrast enhanced brightness correspondence table ^ ^ Two parts contrast enhanced brightness correspondence table includes four different shapes of the song, line 23 · The digital image adaptive color adjustment device described in claim 22 of the patent application, wherein + is to increase the pair of strong highlights: line slip degree 'the contrast enhanced brightness correspondence table generating unit according to the first partial ratio The enhanced brightness correspondence table (LUT_CTF1) is compared with the second portion to enhance the boundary point of the degree correspondence table (LUT_CTF2), and a fine adjustment area is obtained. , Μ Μ 如 如 如 如 如 如 如 如 如 如 如 如 如 如 如 如 如 如 如 如 如 如 如 如 如 如 如 如 如 如 如 如 如 如 如 如 如 如 如 如 如 如 如 如 如 如 如 如(4) 153 料 顚 顚 15 15 15 15 15 15 15 15 15 15 15 15 15 15 15 15 15 15 15 15 15 15 15 15 15 15 15 15 15 15 15 15 15 15 15 15 15 15 15 15 15 15 15 15 15 15 15 15 15 15 15 15 15 15 15 15 15 15 15 15 15 15 15 15 15 15 15 15 15 15 15 15 15 15 15 15 15 15 15 15 15 15 15 15 15 15 15 15 15 15 15 The digital image adaptive color component if= the relationship between the chroma compensation factor and the brightness before and after the adjustment may be a linear relationship or a multiple linear relationship. 7. An image adaptive color adjustment method, comprising: a production wheel person-transfer image signal is converted to continuous-luminance color remaining 5 #u ’ the luminance chrominance separation signal contains-luminance component and chromaticity 28 J2 9 5 ^ ^ j 8631 twf.doc/r component; statistically calculate the probability distribution of the luminance component, and perform adaptive weight distribution to obtain a weight luminance probability distribution map; dynamic range of the weight luminance probability distribution map (DR) extension to pay a dynamic range correspondence table (lut_dr); contrast enhancement of the weight luminance probability map to obtain a contrast enhanced brightness correspondence table (LOT_CT); curve the contrast enhanced brightness correspondence table Smoothing; adding a plurality of dynamic range brightness correspondence tables corresponding to the plurality of brightness components to form a dynamic range brightness correspondence table for one of the subsequent brightness components, and smoothing and comparing the plurality of brightness components corresponding to the plurality of brightness components The enhanced brightness correspondence table is added and averaged to form a contrast enhanced brightness correspondence table applied to the subsequent brightness component; the dynamic range brightness correspondence table and the contrast enhanced 72 degree correspondence table are averaged, and the brightness is performed on the subsequent brightness component Adjusting; according to the brightness component before and after the adjustment, the chrominance component of the luminance chrominance separation signal Saturation compensation line; and the luminance-chrominance signal separation of luminance and saturation adjustment compensation inverse conversion 'to produce a continuous output of the video signal. 28. The method of claim 27, wherein the luminance chrominance separation signal comprises: a YCbCr signal, a Υυν signal, an HSI signal, and a YPbPr signal. The method of claim 27, wherein the step of obtaining the weight probability distribution map comprises: simulating an adaptive visual perception of the human eye when viewing the image 29 1295895 18631 tWf.doc/r, The probability distribution of the luminance component performs luminance distribution weight distribution. 30. The method of claim 27, wherein the step of performing a dynamic range (DR) extension to obtain the dynamic range luminance correspondence table (LUT_DR) comprises: selecting one of the weights of the luminance probability distribution map to be the darkest point The digit value and the most bright point digit value are defined as a first black level point digit value and a white order point digit value, and Ren Siyi is used to extend one of the dynamic range and the second order point digit value and a second white a step value, wherein the second black point value is less than the first black level point value, and the second white point point value is greater than the first white point point value, the second black level point number The value and the second white point point value are used to generate the dynamic range brightness correspondence table. [31] The method of claim 27, wherein the moving bear pulls the vanishing force to the silk generating unit to add g to the weighting brightness probability, and the rate value reaches - the first number of females - the second numerical value of the brightness digit The value comes from the first-black-order Lai Jian and the -the first-self-touch value, (4) 3-=5--the second black-order point digit value and the -second white-order point i-bit value, /, Zhong, Haidiyi The black-order point digit value is smaller than the first black and the second white-order point value is greater than the first-white-order point digit value, the second-order black-order point touch value, and the fresh second white-order point number j-wide brightness correspondence table. If you apply for the third paragraph of the patent scope, or if you want the dynamic range to be bright, you can count on the method of "Red Money". Langsiyi linear or normal interpolation 33. The method described in the 27th item of Shishen Qiao patents, which makes the contrast enhancement of the weight of the party's 30 31,295,895,1,631, ft.doc/r', to obtain the contrast enhanced brightness correspondence table (LUT) The step of -CT) comprises: calculating a dark component probability cumulative value (F1) and a bright component probability cumulative value (F 2 ) of the weight luminance probability distribution, according to the dark component probability cumulative value (F and the bright color The component probability cumulative value (F2), the contrast enhanced brightness correspondence table consists of two sets of parameters covering different ranges of values, the first part of the contrast enhanced brightness correspondence table (LU Ding_CTF1) and the second part contrast enhanced brightness correspondence table (LUT-CTF2) constructed, first The partial contrast enhanced brightness correspondence table and the second partial contrast enhanced brightness correspondence table include four different shapes of curves. 34. The method of claim 33, wherein the step of enhancing the smoothing of the riding curve of the Lianglin County comprises: comparing the brightness enhancement pair (LUT-CTF1) with the first portion The second part compares the boundary point of one of the enhanced party correspondence tables (LUT-CTJF2), and obtains a fine-tuning area to fine-tune the curve smoothness. 35. The method of claim 34, wherein the fine-tuning of the curve smoothness of the fine-tuning region is performed by a sub-linear interpolation method. In the method described in claim 27, the compensation unit is based on the difference between the bright and the slanting parts before and after Na: ς -#度补偿® number to maintain the image The color tone - the sex. ^ 7. The method of claim 27, the difference between the compensation factor and the brightness before and after the adjustment - linear relationship or multiple linear relationship. - The Η can be