Chapter 2 Prior Methods
2.3 Summary
The FSC-LCDs have three times optical throughput than conventional LCDs.
However, the serious issue for the operations is limitation of LC response time. Some auxiliary solutions were proposed for increasing LC response time, such as multi-division backlight technique and overdrive technique. However, these methods cannot provide enough LC response time for large size display. Louis D. Silverstein proposed a spatial-temporal two field method to reduce field number and save longer time for LC response. This two field method still needs color filters which will sacrifice the light throughput. Therefore, we proposed the two-color-field sequential method without color filters which can not only reduce the LC response time but also can promote the optical throughput of FSC-LCDs to 100%.
16
Chapter 3
Two-color-field Sequential Method
Two-color-field sequential method for color filterless LCD was proposed to further reduce the field rate, and provide the longest time for LC response. Thus, many commercial LC modes such as TN, IPS, and MVA modes can achieve the response time. The concept and algorithm of proposed method will present in section 3.1. The following section described the two indexes for evaluating the accuracy of colorimetric reproduction. Finally, the CBU examination will be given.
3.1 Two-color-field Sequential Method
3.1.1 Concept
The concept of proposed two-color-field sequential method is displaying two color-mixing fields with double frame rate to generate a full-color image. The proposed two-color-field sequential method is different from other two-field methods in regard to color filters. The comparison of driving scheme between conventional three field method and proposed two-color-field sequential method is illustrated in Fig.
3-1. The conventional FSC method displays red, green, and blue field images time sequentially with triple frame rate to yield a full-color image. The proposed two-color-filed sequential method flashes two color-mixing fields with two-thirds filed rate of conventional method to generate a full-color image. Two-color-field sequential method uses the least field number, and provides longest time for LC response. Thus, the LC modes can be utilized in commercial LC modes, such as TN,
17
MVA, or IPS. Moreover, the proposed two-color-field sequential method is a type display method for color filterless LCD, so it can take full advantages of optical efficiency enhancement of the temporal color-mixing methods.
B-field G-field
R-field
RB1-field GB2-field Target image
t Field 2
Field 1 Field 3
Field 1 Field 2
Three-field method
Two-color-field method
Fig. 3-1 Two driving-scheme types, a typical three-field method and the proposed two-color-field sequential method, are illustrated, by field decomposition of color fields to generate a full-color image.
3.1.2 LCD Structure
In order to display a color-mixing field in proposed method for LCD without color filters. We propose to incorporate the local color dimming backlight technique [15] to substitute for the equivalent function of the special color filters. The local color dimming backlight technique is a kind of backlight controlling technique, which can locally control LED signals per color per division. This technique was also called high dynamic range (HDR) technique, and the flowchart is illustrated in Fig. 3-2. First of all, normalized target intensity of each pixel was got by input digital LC signals in full on white backlight. Then, LED signals were computed based on target image content by local controlled backlight algorithm, such as maximum, root, or average of signals in each segment [16]. Following, the point spread function was convoluted the LED
18
signals in each segment to get backlight distribution. In step 5, the LC intensity in each pixel was calculated by target intensity divided backlight distribution. Finally, the LED and LC driving signals were got by inversing transfer functions of LED and LCD individually. This algorithm can be used into two different processes, one is the forward process and the other is backward process. The forward process follows at step 3 compute the backlight distribution firstly, and then in step 6 get compensated LC signals. However, sometimes the compensated LC signals will be got firstly, thus the process will be inversed to calculate LED driving signals. Using the local dimming backlight system can dim or boost backlight intensity in each segment depend on image content. Thus, the HDR technique has low power consumption, high contrast ratio, and high color saturation advantages. Moreover, the local controlled backlight technique can be utilized to achieve color-mixing fields of the two-color-field sequential method.
I 1
I : target intensity of single pixel (normalized) r1: transfer function of LED
p1: point spread function r2: transfer function of LCD
Forward: Step 1⇒ Step 3⇒ Step 6 Backward : Step 1⇒ Step 6⇒ Step 3
Fig. 3-2 The local dimming system flowchart.
19
3.1.3 Algorithm
The procedure starts at transforming the digits of an input image into the corresponding tri-stimulus values of each primary channel as the target information.
The transformation is accomplished through a suitable color model [17], as detailed in the next section. Then, the red and the green, for example, are chosen as the first and the second primaries (Fig. 3-3). After being dealt with by the forward process of local color dimming backlight technique, the red information in the first field is achieved by the sets of LED signals (dBL-R) and LC signals (dLC1), while the green one in the second field by the sets of dBL-G and dLC2. The forward process features the sequence of the first deriving the light-emitting diode (LED) signals and then computing the LC compensation signals to achieve the target information; the backward process reverses the forward one [18].
As for the blue one in the first field, the blue target information is dealt with via the backward process, based on dLC1, to deduce the blue LED signals (dBL-B1) for the first field. In general, the blue information, resulted from the combination of dLC1 and dBL-B1, may be different from the blue target information. Therefore, the difference between the original and the resulted blue information in the first field is set as the new blue target information for the second field. Following the same procedure, the set of blue LED signals (dBL-B2) for the second field is obtained.
In practice, the accuracy of the third color reproduction depends on backlight layout, light spread function (LSF), and image content. The algorithm should be iterated with additional optimization process to improve color presentation accuracy. In addition, the decision of the third primary is not unique. The blue information, in the aforementioned example, is selected since the human vision system is less sensitive to blue information. Besides, a primary with least significant content is a useful option to
20
increase reproduction accuracy.
1stfield image 2ndfield image
Target image Rdimming B/L
(dBL-R) Gdimming B/L
(dBL-G)
+ +
Local dimming algorithm Local dimming algorithm
(dBL-B2 ) (dBL-B1 )
B1dimming B/L B2dimming B/L
1stfield LC(dLC1 ) 2ndfield LC(dLC2 )
2ndfield B/L 1stfield B/L
forward process reverse process
Fig. 3-3 The algorithm flowchart of the two-color-field sequential method.
3.2 Colorimetric Reproduction
In order to predict the accurate color information in display system, the colorimetric characterizations of displays were determined. In 1998, Fairchild and Wyble, recognizing the fundamental differences between liquid-crystal and cathode-ray-tube technologies, develop a successfully LCD color model [19]. The color model configuration is described in Eq.3-1. The process is divided into two main stages. The first non-linear stage of the model is built up three one-dimensional look-up tables (LUTs) of the radio-scales in each channel, as shown in Eq.3-1-1. The LUTs described the relationship between the digital input signals used to drive a display and the radiant output produced through LC cells by accounting for non-linear optoelectronic transfer function, OETF. Where, d defines digital counts and R, G, and B are radio-scales for red, green, and blue channels, respectively. The radio-scales’
21
ranges were conformed from 0 to 1. The second linear stage, as shown in Eq.3-1-2, represented the correlations between radio-metric scalars and the resultant tri-stimulus values. The flare term, [Xk, Yk, Zk]T accounts for the radiant output at black level in LCDs, since the liquid crystal having a minimum transmittance factor above zero.
Moreover, the “max” subscript defines each channel’s maximum output and subscript
“kmin” defines the black-level radiant output. Furthermore, this color model performs non-linear optimization to minimize the mean CIEDE2000 color difference of test colors sampling the display’s gamut in complex models. However, this color model is not suitable used in color dimming backlight system.
0 ≤ R,G,B ≤ 1
Therefore, the color model should be modified to incorporate the color backlight intensity. Then, the equation was transformed, as shown in Eq.3-2 [17], where, L is the normalized backlight intensity in each color which can be obtained by the convolution computation with LED signals and light spread function of each LED.
The subscript “r, g, and b” defines red, green, and blue colors respectively. Summarily, this color model was suitable to be used for predicting color information in our proposed system.
22
3.2.1 Color Difference Formula
Color difference formula is used to evaluate the color difference between two static images. In our research, the color difference index would be used to verify the colorimetric reproduction accuracy. Since 1976, the International Commission on Illumination (CIE) recommended two color-difference formula for industrial applications, the CIELAB and CIELUV formula [20]. The modification formula, CIEDE2000 [20], includes not only lightness, chroma, and hue weighting functions, but also an interactive term between chroma and hue differences for improving the performance for CIELAB color difference indexes. There are four steps include in CIEDE2000 calculation. In the first step, calculate the CIELAB as shown in Eq.3-3-1, the parameters L＊ represents lightness, a＊ approximate redness-greenness, b＊ approximate yellowness-blueness, and C＊ab chroma. Then, compute a’, C’, and h’, follow Eq.3-3-2, in this step, bar C＊ab is the arithmetic mean of the C＊ab values for a pair of samples. The third step, obtain ΔL’, ΔC’ and ΔH’ values between standard and sample in a pair as shown in Eq.3-3-3. At last step, calculate the color difference values using CIEDE2000 formula in Eq.3-3-4. The parameters SL, SC, and SH are the
23
weighting functions for lightness, chroma, and hue differences, respectively. kL, kC, and kH values are the parametric factors to be adjusted according to different viewing parameters, for the lightness, chroma, and hue components, individually. RT function is intended to improve the performance of color-difference equation for fitting chromatic differences in the blue region. The color difference formula CIEDE2000 considers more color conditions, and it will be more suitable for evaluating the color differences.
(Eq.3-3-1)
(Eq.3-3-2)
24
(Eq.3-3-3)
(Eq.3-3-4)
3.2.2 Spatial-CIELAB (S-CIELAB)
The CIE color difference formulae are developed to measure the color difference between color patches with small color difference after moderate chromatic adaptation.
However, this value does not give satisfactory results in human visual system;
because of the point-by-point computation result in complex image is always larger than observer’s visibility. Therefore, X.Zhang proposed extension of the CIELAB color metric, Spatial-CIELAB(S-CIELAB) to measure color reproduction errors in images [21]. The Spatial-CIELAB flowchart is illustrated in Fig. 3-4. In the first step, transform the input images into a device independent space, CIE 1931 XYZ tri-stimulus values. The second step, put tri-stimulus values into opponent-color space, AC1C2.These channels were determined through series psychophysical experiments testing for pattern-color separability [22], where A denotes the luminance channel and C1, C2 are chrominance channels. The opponent channels are a linear transform CIE
25
1931 XYZ as shown in Eq.3-4. The third step, operate the opponent sensitivity signals to frequency domain by Fourier transformation. These three independent channels can be spatially filtered, using filters that approximate the contrast sensitivity function (CSF) of the human visual system. Each channel can accomplish using multiplications in the frequency domain. Moreover, a three parameter exponential model, described by Movshon, is description of the general shape of luminance CSF, which model is shown in Eq.3-5, where, f is the spatial frequency in unit of cycle per degree, the parameters, a, b, c were fit to existing experimental data, and the normalized frequency filter for luminance channel as shown in Fig. 3-5.The CSF of other two chrominance channels is expressed by Eq.3-6, and the normalized opponent color contrast sensitivity is described in Fig. 3-6. The available data can be fitted with the sum of Gaussian functions, and the parameters for chrominance CSF that were fitted to the Van der Horst and Poirson data sets is shown in Table. 3. Finally, sum up the difference between target image and reproduced image computed by CIELAB color difference formula, the output value were evaluated the colorimetric reproduction errors. Consequently, the S-CIELAB difference measure reflects both spatial and color sensitivity, due to psychophysical experiments verified, the index is more suitable to evaluate the image reproduction errors.
26
Color image
Color separation
Lum R/G B/Y
Spatial filter
X Y Z
S-CIELAB
Fig. 3-4 The S-CIELAB floechart
(Eq.3-4)
(Eq.3-5)
Fig. 3-5 The normalized luminance contrast sensitivity and frequency filter for luminance channel.
27
(Eq.3-6)
Table. 3 Parameters for chrominance CSFs
Fig. 3-6 The normalized opponent color contrast sensitivity
Fig. 3-7 The frequency filters for chrominance channels.
28
3.3 Color Break-up Examination
A relative contrast sensitivity (RCS) index was used to evaluate the color break-up phenomenon [23]. The RCS values were computed following the procedures in the S-CIELAB with some modifications, as shown in Fig. 3-8. Firstly, transform the digits of an input image and CBU image into the corresponding tri-stimulus values of each primary channel. Then, tri-stimulus values, XYZ transform into opponent-color space, AC1C2, where A denotes the luminance-related channel signals; C1 as the red-green channel signals and C2 is the blue-yellow channel signals. Third, each channel of opponent-color signals convolute the contrast sensitivity function (CSF) to simulate the spatial blurring by the human visual system. Forth, subtract the filtered channel images between each channel to extract the CBU fringe. Finally, the subtracted result of each pair of filtered channel images are summed together in different weights. The outcome is relative contrast sensitivity, RCS. The CBU phenomenon will be evaluated by RCS index to reflect the stimulus of CBU fringes in human visual system.
29
Fig. 3-8 The RCS flowchart
3.4 Summary
The two-color-field sequential method proposed to further reduce the field number which can result in longer time for LC response, which commercial LC modes can achieve. In this method, we incorporate local color dimming backlight technique to substitute for the equivalent function of the special color filters, and keep the whole optical throughput.
The S-CIEDE2000 index combined a spatial modulation of human visual system before calculating CIEDE2000 values. The RCS value reflects humans’ response for CBU edge stimulus. Therefore, the two indexes, S-CIEDE2000 and RCS will be used to verify the experimental accuracy in follow sections.
30
Chapter 4
Experimental Results
A verification of color reproduction accuracy was based on a 37” HDR display with 8 by 8 locally controlled color backlight divisions. The light spread function (LSF) of each division and display characteristic were measured to predict the LED and LC driving signals. Color difference values between the target image and the reproduced image were evaluated for color reproduction accuracy. Moreover, CBU examination will be presented in follow sections.
4.1 Experiments
The experimental apparatus are illustrated in Fig. 4-1. The LED and the compensated LC signals were calculated by using computer. The driving signals input to the HDR panel, and then the colorimeter measured tri-stimulus. The experimental flowchart is shown in Fig. 4-2.
First, the light spread functions per color per divisions were measured since the HDR display is a spatial shift-variant system, as shown in Fig. 4-3. Each LSF had different profile, especially at the display edges. Therefore, the measurement of light spread function was used to predict accurate LED signals. Secondly, LCD characteristics were measured including tri-stimulus values of maximums and flare terms in each primary color, also the OETF of LC cells in the panel. Then, measured parameters were used to establish a suitable color model for this panel, as mentioned in section 3.2. The following step, depend on a two-color-field algorithm to predict the LED and LC signals. Next step, input the calculated signals into the panel, and
31
measured the tri-stimulus values of the test image. Finally, calculate the color difference between target image and reproduced image. Therefore, color difference values evaluated the accuracy of color reproduction.
37” HDR display with 64 BL segments
Colorimeter Computer
Fig. 4-1 The experimental apparatus.
Measurement of LSF in each B/L segment
Measurement of display characteristic Establishment of color model
Two-color-field sequential method
Pure color verified Spatial frequency variation
Fig. 4-2 The experimental flowchart.
32
Fig. 4-3 The spatial shift-variant system.
4.1.1 Light Spread Function Measurement
The experimental results of light spread function measurement are illustrated in Fig. 4-4. In the two-color-field sequential method algorithm, the backlight distribution was gathered using the superposition method. Comparing the conventional convolution method with the superposition method, the convolution method measured one light spread function to simulate each light spread function and used convolution algorithm to get backlight distributions. On the other hand, the superposition method measures every light spread function and adds each light distribution to obtain backlight distribution. For example, in the full on white backlight; the superposition method has a higher correlation coefficient of 95% with measurement result than convolution methods of 83% as shown in Fig. 4-5. Therefore, the light spread function of 64 backlight divisions of three primary colors were measured to predict accurate LED signals and compensated LC signals.
33
Fig. 4-4 The experimental results of light spread function measurement.
83% 95%
(a) (b) (c) Fig. 4-5 The comparison of (a) measurement full on white backlight and (b) the convolution method and (c) the superposition method.
34
4.1.2 Panel Characteristic Measurement
In order to obtain device-independent color, the color model was established. As section 3.2 mentioned, a suitable color model needs three 3-dimensional look-up tables and the tri-stimulus of the panel. Therefore, 37” HDR display parameters were measured to build the color model.
When designing the look-up tables, the LC driving signals were divided ten parts from 0 to 255 gray levels in full-on backlight. Measuring the tri-stimulus, X, Y, Z, of these eleven levels, and the correlation between the digital input signals used to drive a display and the radiant output produced through LC cells by accounting for non-linear optoelectronic transfer functions were obtained. The 3-dimensional look-up tables of the three primary colors are illustrated in Fig. 4-6.
In the second part of the measurement process, the maximum and flare terms of tri-stimulus values in the HDR panel were measured by using the colorimeter. The maximums of tri-stimulus were measured with full-on backlight and largest LC driving signals in each primary color individually. Thus, Eq.3-2 was transformed to Eq.4-1. Consequentially, the result color model was used to predict the accurate LED and compensated LC signals.
35
Fig. 4-6 Three 3-dimentional LUTs of (a) red, (b) green, and (c) blue.
r
36
4.2 Results
The results of colorimetric reproductions are illustrated in 4.2.1. The test images include pure color images with different gray levels and random color patch images with differing spatial frequency. Moreover, the results of color break-up examinations will be given.
4.2.1 Color Differences
The color reproduction verifications include two parts, pure color image reproductions and the color patch images reproductions. The twelve pure color images of 10, 90, 180, and 250 gray levels in three primary colors were chosen as the test images. Moreover, the average CIEDE2000 values evaluated the accuracy of
The color reproduction verifications include two parts, pure color image reproductions and the color patch images reproductions. The twelve pure color images of 10, 90, 180, and 250 gray levels in three primary colors were chosen as the test images. Moreover, the average CIEDE2000 values evaluated the accuracy of