A multi-angle color correction method based on cylindrical grating three-dimensional display

Abstract

The application discloses a multi-angle color correction method based on a cylindrical grating three-dimensional display, which comprises the following steps: displaying a test image; performing spectral measurement on the test image, and correcting a hardware system according to a test result; collecting data at different angles to obtain color tristimulus values XYZ; establishing a color characteristic model for each angle according to different spectral measurement angles in the third step; establishing a full-angle model; performing color space conversion on a to-be-corrected image, converting the to-be-corrected image to a device-independent standard color space, and storing XYZ information of the to-be-corrected image; performing color correction at different angles; and reproducing the image. The multi-angle color correction method can complete color correction of a display stereoscopic image through multi-angle color characteristic measurement.

Description

Technical Field

[0001] This invention relates to the field of image processing technology, and more specifically to a multi-angle color correction method based on cylindrical grating three-dimensional display. Background Technology

[0002] Cylindrical lenticular lens 3D display technology is an important stereoscopic display technology. It achieves naked-eye stereoscopic display by adding cylindrical lenticular lenses to a flat-panel display, creating the effect of different images seen by the left and right eyes. Compared to other stereoscopic display technologies, cylindrical lenticular lens 3D display technology offers excellent observability, realistic display effects, and a wide range of applications, including virtual reality, gaming, medicine, industrial design, and education. When using cylindrical lenticular lens 3D display technology, the displayed colors significantly impact the viewing experience. Inaccurate color display greatly reduces the realism and stability of the stereoscopic effect.

[0003] Currently, color calibration for ordinary displays can be achieved through color characterization. Color characterization is a technique used to measure and analyze the color characteristics and color space of a device, converting the device's color performance into a standard color space and color characteristics, thereby enabling accurate control and management of color performance. However, cylindrical lenticular 3D display technology displays multi-angle stereoscopic images, and the simple color characterization process cannot meet the color calibration requirements for stereoscopic displays.

[0004] Therefore, it is necessary to combine color characterization technology to measure the color characteristics of images displayed on 3D display devices from various angles, and perform multi-angle color correction based on different color characterization models for each angle. On this basis, the corrected color information of each pixel at each angle is used for 3D spatial encoding to obtain a color-accurate 3D image display. Summary of the Invention

[0005] This invention provides a multi-angle color correction method based on cylindrical grating three-dimensional display. This correction method can complete the color correction of the displayed stereoscopic image through multi-angle color characteristic measurement.

[0006] The present invention adopts the following specific technical solution:

[0007] A multi-angle color correction method based on cylindrical grating 3D display, the multi-angle color correction method includes the following steps:

[0008] The first step is to display the test images. Two test images are set: one is a pure white image with RGB values ​​(255, 255, 255), and the other is an RGB image divided into 120 regions, each region being a pure color region with different RGB values. After 3D encoding of the two test images, they are displayed using a cylindrical raster 3D display device. When 3D encoding the RGB images, each pure color region uses the same RGB values ​​at all angles, while the RGB values ​​between different pure color regions are different.

[0009] The second step is performance testing, which involves spectral measurement of the test images to test the spatial uniformity, temporal stability, and channel additivity of the cylindrical grating 3D display device, and making corrections to the hardware system based on the test results.

[0010] The third step is to collect data from multiple angles. Using a spectral measurement device, the three-dimensional image displayed in the first step is measured from multiple angles within the range of 0° to 180°. The tristimulus values ​​XYZ can be obtained by integrating the measured spectra.

[0011] The fourth step is to establish a color characteristic model from different angles. For different spectral measurement angles in the third step, a color characteristic model is established for each angle. At each angle, the RGB values ​​of different pure color areas displayed in the first step are combined with the corresponding XYZ values ​​measured in the third step to establish a mapping relationship from the color characteristics RGB of the device-dependent space to the color characteristics XYZ of the device-independent standard space, which is the color characteristic model from different angles.

[0012] The fifth step is to establish a full-angle model. The color characteristic model obtained in the fourth step is interpolated to complete the characteristic data of each observation angle other than the sampling angle, so as to obtain a full-angle color characteristic model for full-angle observation within 0° to 180°.

[0013] The sixth step is color space conversion. The image to be corrected is converted to a device-independent standard color space, and its XYZ information is stored.

[0014] The seventh step is to perform color correction from different angles. At each angle, the full-angle color characteristic model established in the fifth step is applied. Using the XYZ information stored in the sixth step as the standard, the RGB values ​​of each pixel in the image to be corrected at each angle are calculated in reverse to obtain the corrected three-dimensional display encoding information.

[0015] The eighth step is image reproduction. The RGB information of the image pixels obtained in the seventh step is used to perform three-dimensional display encoding, and the corrected image is displayed on a cylindrical grating three-dimensional display device.

[0016] Furthermore, in the first step, the setting of 120 RGB pure color areas needs to have both single-channel driven and multi-channel driven images, specifically set as follows:

[0017] Within the single-channel range of 15-255, R, G, and B single-channel driving images are set at intervals of 15; within the single-channel range of 15-255, RG combination, RB combination, and GB combination dual-channel driving images are set at intervals of 15, ensuring that the channel driving values ​​within the combination are equal; within the single-channel range of 0-255, RGB combination three-channel driving images are set at intervals of 15, ensuring that the driving values ​​of the RGB three channels in each combination are equal, resulting in 18 combinations from (0,0,0) to (255,255,255).

[0018] Furthermore, in the third step, the specific settings for multi-angle spectral measurement are as follows: within the range of 0° to 180° with respect to the 3D display panel, a spectral measurement is performed every 5° interval, that is, a total of 31 measurements are performed for each image.

[0019] Furthermore, in the fourth step, for each angle, the mapping relationship is established using one of the following methods: three-dimensional lookup table (LUT), matrix transformation, piecewise linear interpolation assuming constant chromaticity (PLCC), piecewise linear interpolation assuming varying chromaticity (PLVC), and gain-bias-gamma model (GOG).

[0020] Furthermore, in the fifth step, for other angles not at the sampling angle, the color characterization model needs to be interpolated using the characterization models of the sampling angles on both sides.

[0021] Furthermore, in the fifth step, the interpolation method is either linear interpolation or polynomial interpolation.

[0022] Furthermore, in the sixth step, the color space conversion of the image to be displayed is accomplished by converting the color information of each pixel, and is achieved through the conversion matrix of different color spaces.

[0023] Furthermore, in the seventh step, at each angle, the XYZ value of each pixel needs to be converted to RGB value using the color characteristic model corresponding to that angle, so as to obtain the RGB information of each pixel at all angles in the range of 0° to 180°, and use the complete RGB information to perform the three-dimensional display encoding in the eighth step.

[0024] Furthermore, the spectral measurement device is a spectrometer, a multispectral camera, a spectrophotometer, or a colorimeter.

[0025] Beneficial effects:

[0026] The multi-angle color correction method of this invention includes displaying a test image, performance testing, acquiring data from different angles, establishing a color characteristic model from different angles, establishing a full-angle model, color space conversion, performing color correction from different angles, and image reproduction. This correction method uses color characteristic technology to measure and analyze the color characteristics of a cylindrical grating 3D display device at a series of angles. Based on the measurement results, an full-angle color characteristic model is obtained through interpolation. On this basis, the color information of each pixel in the displayed image at each angle is adjusted, and 3D spatial encoding is performed to achieve the desired color correction effect. Through these steps, the function of multi-angle color correction for a cylindrical grating 3D display device is realized. It can accurately measure and analyze the color characteristics of the 3D display device from various angles, thereby correcting the color of the 3D display device and improving the consistency and accuracy of color when observing 3D images from various angles. Attached Figure Description

[0027] Figure 1 This is a flowchart of the multi-angle color correction method of the present invention;

[0028] Figure 2 A schematic diagram of a test image divided into 120 regions;

[0029] Figure 3 The specific location for performing spatial uniformity testing on the test image;

[0030] Figure 4 This is a schematic diagram showing the specific angles used during spectral measurements.

[0031] Figure 5 A flowchart for establishing a color characterization model using the PLCC method;

[0032] Figure 6 This is a flowchart for calculating the driving value of the color characteristic model established using the PLCC method. Detailed Implementation

[0033] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0034] Example 1

[0035] like Figure 1 As shown, this embodiment provides a multi-angle color correction method based on cylindrical grating three-dimensional display. The multi-angle color correction method includes the following steps:

[0036] The first step, S1, involves displaying the test images. Two test images are set: one is a pure white image with RGB values ​​(255, 255, 255), and the other is an RGB image divided into 120 regions. Figure 2 As shown, each region is a pure color region with different RGB values, that is, 120 regions have 120 different RGB values; after three-dimensional encoding of the two test images, they are displayed using a cylindrical grating three-dimensional display device; when encoding the RGB images for three-dimensional display, each pure color region uses the same RGB value at all angles, and the RGB values ​​between different pure color regions are different;

[0037] The setup for 120 RGB solid color areas requires both single-channel and multi-channel driven images. Specifically, within the single-channel range of 15-255, set R, G, and B single-channel driven images at intervals of 15. Taking the R channel as an example, there are 17 combinations from (15,0,0) to (255,0,0). Within the single-channel range of 15-255, set RG, RB, and GB dual-channel driven images at intervals of 15, ensuring that the channel driving values ​​within each combination are equal. For example, one possible setting for the RG combination is (15,15,0). Within the single-channel range of 0-255, set RGB three-channel driven images at intervals of 15, ensuring that the driving values ​​of the three RGB channels in each combination are equal, resulting in 18 combinations from (0,0,0) to (255,255,255).

[0038] The second step, S2, is performance testing. Spectral measurements are performed on the two test images to test the spatial uniformity, temporal stability, and channel additivity of the cylindrical grating 3D display device. Based on the test results, the hardware system is corrected.

[0039] The third step, S3, involves multi-angle data acquisition. A spectral measurement device is used to perform multi-angle spectral measurements on the three-dimensional image displayed in the first step, ranging from 0° to 180°. Integrating the measured spectra yields the tristimulus values ​​XYZ. Figure 3 As shown, the specific settings for multi-angle spectral measurement are as follows: within the range of 0° to 180° with respect to the 3D display panel, a spectral measurement is performed every 5° interval, that is, a total of 31 measurements are performed for each image; the spectral measurement equipment can be a spectrometer, a multispectral camera, a spectrophotometer, or a colorimeter, etc.

[0040] The fourth step, S4, involves establishing a color characteristic model for each angle. For each different spectral measurement angle from the third step, a color characteristic model is established. At each angle, the RGB values ​​of the different pure color regions displayed in the first step are combined with the corresponding XYZ values ​​measured in the third step to establish a mapping relationship from the color characteristics RGB of the device-dependent space to the color characteristics XYZ of the device-independent standard space—this is the angle-specific color characteristic model. For each angle, the mapping relationship is established using methods such as the three-dimensional lookup table (LUT), matrix transformation, piecewise linear interpolation assuming constant chromaticity (PLCC), piecewise linear interpolation assuming varying chromaticity (PLVC), or gain-bias-gamma model (GOG).

[0041] Step S5: Establish a full-angle model. Perform interpolation on the sub-angle color characteristic model obtained in step four to complete the characteristic data of each observation angle other than the sampling angle, and obtain a full-angle color characteristic model for full-angle observation within 0° to 180°. For other angles not on the sampling angle, the color characteristic model needs to be interpolated using the characteristic models of the sampling angles on both sides. The interpolation method can be linear interpolation or polynomial interpolation.

[0042] Step S6: Color space conversion. The image to be corrected is converted to a device-independent standard color space, and its XYZ information is stored. The color space conversion of the image to be displayed is accomplished by converting the color information of each pixel, and is achieved through the conversion matrix of different color spaces.

[0043] Step S7: Perform color correction by angle. At each angle, apply the full-angle color characteristic model established in Step 5, using the XYZ information stored in Step 6 as the standard, and inversely calculate the RGB value of each pixel in the image to be corrected at each angle to obtain the corrected 3D display encoding information. At each angle, the XYZ value of each pixel needs to be inversely calculated to the RGB value using the color characteristic model corresponding to this angle, to obtain the RGB information of each pixel at all angles within the range of 0° to 180°. Use the complete RGB information to perform 3D display encoding in Step 8.

[0044] Step S8: Image reproduction. The RGB information of the image pixels obtained in step 7 is used to perform 3D display encoding at various angles, and the corrected image is displayed on a cylindrical grating 3D display device.

[0045] The aforementioned multi-angle color correction method includes displaying test images, performance testing, multi-angle data acquisition, establishing color characteristic models for each angle, establishing a full-angle model, color space conversion, multi-angle color correction, and image reproduction. This correction method uses color characteristic technology to measure and analyze the color characteristics of the cylindrical grating 3D display device at a series of angles. Based on the measurement results, an full-angle color characteristic model is obtained through interpolation. On this basis, the color information of each pixel in the displayed image at each angle is adjusted, and 3D spatial encoding is performed to achieve the desired color correction effect. Through these steps, the function of multi-angle color correction for the cylindrical grating 3D display device is realized. It can accurately measure and analyze the color characteristics of the 3D display device at various angles, thereby correcting the color of the 3D display device and improving the consistency and accuracy of color when observing 3D images from various angles.

[0046] LUT: Lookup Table, a method used for color characterization. It uses interpolation and extrapolation algorithms to directly obtain the bidirectional correspondence between XYZ and RGB. Points falling on the sampling grid vertices take the vertex value, while points not on vertices use the nearest surrounding vertex obtained through an interpolation algorithm.

[0047] PLCC: Piecewise linear interpolation with constant chromaticity, a method for color characterization. Due to the assumption of constant chromaticity, the chromaticity coordinates at the maximum value of a single channel can be used instead of the chromaticity coordinates to be solved. Only the luminance is interpolated, and the final result is obtained by combining the chromaticity coordinates and luminance values.

[0048] PLVC: Piecewise linear interpolation assuming chromaticity variation, a method for color characterization. Considering that the chromaticity coordinates of a displayed image may not necessarily satisfy the assumption of constant chromaticity coordinates, both a luminance lookup table and a chromaticity lookup table are established. In practice, the luminance value is first determined using the luminance lookup table based on the driving value, and then the chromaticity coordinates are determined using the chromaticity lookup table based on the luminance value. Finally, the chromaticity coordinates and luminance values ​​are combined to obtain the result.

[0049] In one specific implementation, the second step of the above-described multi-angle color correction method can be achieved through the following process:

[0050] First, the spatial uniformity of the 3D display device is tested. For example... Figure 3 As shown, using the pure white test image set in step S1, the image area was divided into 9 equal parts. Spectral measurements were performed on the central region of each part, and the results were recorded as X1Y1Z1-X9Y9Z9. These were then converted to the CIE1976 u'v' space for color difference calculation. The result, less than 0.0033jnd, demonstrates good spatial uniformity. The average color difference between the 9 regions and the central region was calculated using the following formula:

[0051]

[0052] In the above formula, DE is the average color difference between the 9 regions and the center region, u'v' is the chromaticity coordinate of the measured color in the CIE1976 u'v' space; i is 1, 2, 3...9, representing the i-th region among the 9 regions.

[0053] Secondly, the temporal stability of the 3D display device was tested. Using the pure white test image set in step S1, multiple spectral measurements were performed on a single point in its central region, with each measurement spaced 30 minutes apart, for a total duration of 5 hours. The 11 measurement results were recorded as X1Y1Z1-X 11 Y 11 Z 11 The color difference was calculated by converting the color space to the CIE 1976 u'v' color space. The result was less than 0.0033jnd, indicating good time stability. The average color difference between the 11 measurements and the first measurement can be calculated using the following formula:

[0054]

[0055] In the above formula, DE is the average color difference between the 11 measurements and the first measurement, u'v' is the chromaticity coordinate of the measured color in the CIE1976 u'v' space, and i is 1, 2, 3... 11, representing the i-th measurement among the 11 measurements.

[0056] Then, the channel additivity of the 3D display device is tested. For example... Figure 2 As shown, using the RGB image set in step S1, spectral measurements are performed at the same angle in four pure color regions with RGB values ​​of (255,0,0), (0,255,0), (0,0,255), and (255,255,255). The results are recorded as X1Y1Z1-X4Y4Z4, respectively. The sum of the tristimulus values ​​of the single-channel image is calculated using the following formula. If it is substantially consistent with the tristimulus values ​​of the three-channel image, it proves that the device satisfies channel additivity.

[0057]

[0058] In the above formula, XYZ are the color tristimulus values ​​obtained by measuring the spectral integral.

[0059] S204: Correct the hardware system of the cylindrical grating 3D display device based on the test results.

[0060] like Figure 5 As shown, in the above multi-angle color correction method, the specific implementation process of the fourth step is illustrated using the PLCC method for color characterization as an example:

[0061] First, for each measurement angle in the third step, record the color characteristic data at this angle, denoted as XYZ(θ), where θ is the measurement angle.

[0062] Secondly, in XYZ(θ), the tristimulus values ​​of the regions with the maximum single-channel driving value, namely the three pure color regions with RGB values ​​of (255,0,0), (0,255,0), and (0,0,255), are denoted as X. R Y R Z R X G Y G Z G X B Y B Z B The chromaticity coordinates x are calculated using the following formulas. R y R z R x G y G z G x B y B z B :

[0063]

[0064] In the formula, XYZ represents the tristimulus values ​​of color, and xyz represents the chromaticity coordinates.

[0065] Next, change the chromaticity coordinate x R y R z R x G y G z G x B y B z B After normalizing the brightness, we obtain the transformation matrix A(θ) for this angle:

[0066]

[0067] Then, based on the measured spectra of all regions in XYZ(θ), three one-dimensional lookup tables, denoted as LUT(θ), are established for the three RGB channels at this angle, so that brightness prediction can be performed from the input digital driving value.

[0068] Finally, repeat the above steps for the measurement data of all sampling angles to establish a color characteristic model for each sampling angle.

[0069] Furthermore, the brightness lookup table established in step S404 can be converted into a "brightness-to-RGB-drive" curve form to perform interpolation of the full-angle model and calculation of brightness interpolation.

[0070] In the above multi-angle color correction method, the specific implementation process of the fifth step is as follows:

[0071] First, to obtain the characteristic matrix at a specific angle, linear interpolation is performed on the characteristic matrices of the two sampling angles closest to this angle on either side of it. This angle is denoted as θ. i The sampling points on both sides are θ i-1 and θ i+1 Then the transformation matrix A(θ) i )for:

[0072]

[0073] Secondly, for angle θ i The RGB three-channel brightness lookup table is also based on θ. i-1 and θ i+1 Linear interpolation is performed using a lookup table, which can then be applied to the brightness curve to obtain the result for angle θ. i The brightness curve corresponds to the angle θ. i Channel brightness lookup table LUT(θ) i ).

[0074] Then, the resulting transformation matrix A(θ) i ) and brightness lookup table LUT(θ) i By combining these, we obtain θ. i Color characteristic model at different angles.

[0075] Finally, within the range of 0°-180°, the above process is repeated for each angle to obtain a full-angle color characteristic model.

[0076] like Figure 6 As shown, in the above multi-angle color correction method, the seventh step S7 is an implementation of color correction for the PLCC color characterization method, and the specific steps are as follows:

[0077] First, at a specific angle, the XYZ information of the displayed image stored in step S6 is combined with the transformation matrix A(θ) of the full-angle color characteristic model in step S5 at this angle to calculate the required single-channel brightness Tr(θ), Tg(θ), and Tb(θ) for the three RGB channels at this angle. The following formula can be used for calculation:

[0078]

[0079] In the formula, XYZ represent the tristimulus values ​​of the displayed image.

[0080] Then, based on the three-channel brightness lookup table LUT(θ) corresponding to this angle stored in step S5, or the "brightness corresponds to RGB drive" curve, the single-channel brightness values ​​Tr(θ), Tg(θ), and Tb(θ) are substituted and interpolated to obtain the required drive value RGB(θ) for this angle.

[0081] Finally, within the range of 0° to 180°, repeat the above steps for all angles required for 3D encoding to obtain the RGB driving information required for each pixel of the displayed image at all angles.

[0082] Example 2

[0083] This embodiment provides a multi-angle color correction method based on cylindrical grating three-dimensional display, which specifically includes the following steps:

[0084] Step S1: Display the test images. Two test images are set: one is a pure white image with RGB values ​​of (255, 255, 255); the other is an RGB image divided into 120 regions, each region being a pure color image with different RGB values. After 3D encoding of both test images, they are displayed using a cylindrical raster 3D display device. The pure color images of different regions of the RGB image can be set as follows: within the range of 15 to 255, with intervals of 15, set R, G, B single-channel driven images; within the single-channel range of 15 to 255, with intervals of 15, set RG combination, RB combination, and GB combination dual-channel driven images, ensuring that the channel drive values ​​within the combination are equal; within the single-channel range of 0 to 255, with intervals of 15, set RGB combination three-channel driven images. During the 3D display encoding of the RGB image, each pure color region uses the same RGB values ​​at all angles, while the RGB values ​​differ between different pure color regions.

[0085] Step S2: Performance Testing. By performing spectral measurements on specific areas of the two test images from Step S1, the spatial uniformity, temporal stability, and channel additivity of the cylindrical grating 3D display device are tested. Based on the test results, hardware system modifications are made, specifically including:

[0086] The spatial uniformity of 3D display devices is tested. For example... Figure 3 As shown, using the pure white test image set in the first step S1, the image area is divided into 9 equal parts, and spectral measurements are performed on the central area of ​​each part. The colorimetric values ​​of the 9 measurements are then analyzed.

[0087] The temporal stability of the 3D display device was tested. Using the pure white test image set in step S1, multiple spectral measurements were taken at 30-minute intervals for a total of 5 hours at a single point in its central region. The chromaticity values ​​of the 11 measurements were then analyzed.

[0088] The channel additivity of 3D display devices was tested. For example... Figure 3 As shown, using the four specific pure color regions of the RGB image set in the first step S1, spectral measurements are performed at the same angle, and the chromaticity values ​​of the four measurements are analyzed.

[0089] The hardware system of the cylindrical grating 3D display device was modified based on the test results.

[0090] Step S3: Angle-by-angle data acquisition. For example... Figure 4 As shown, a spectrometer is used to perform multi-angle spectral measurements on the three-dimensional image displayed in the first step S1, ranging from 0° to 180° with 5° intervals. The measured spectrum represents the light intensity at a specific wavelength within a certain band, and integrating these measurements yields the tristimulus values ​​XYZ.

[0091] Step S4: Establish a color characteristic model from different angles. For example... Figure 5 As shown, for different spectral measurement angles in step S3, a color characterization model is established for each angle using the PLCC method. At each angle, the RGB values ​​of different pure color regions displayed in step S1 are combined with the corresponding XYZ values ​​measured in step S3. The PLCC method is then used to establish a mapping relationship from the device-dependent color characteristics RGB to the device-independent standard color characteristics XYZ, i.e., the color characterization model. This specifically includes:

[0092] For each measurement angle in step S3, the color characteristic data at this angle is recorded and denoted as XYZ(θ). In XYZ(θ), the tristimulus values ​​of the regions with the maximum single-channel driving value, i.e., the three pure color regions with RGB values ​​of (255,0,0), (0,255,0), and (0,0,255), are denoted as X. R Y R Z R X G Y G Z G X B Y B Z B Calculate their chromaticity coordinates x and y respectively. R y R z R x G y G z G x B yB z B . Chromaticity coordinates x R y R z R x G y G z G x B y B z B Brightness normalization is performed to obtain the transformation matrix A(θ) for this angle; based on the measured spectrum of all regions in XYZ(θ), three one-dimensional lookup tables, denoted as LUT(θ), are established for the three RGB channels at this angle, so that brightness prediction can be performed from the input digital driving value; the above process is repeated for the measurement data of all sampling angles to establish a color characteristic model for each sampling angle.

[0093] Step 5 (S5): Establish a full-angle model. Interpolate the segmented-angle color characteristic model obtained in Step 4 (S4) to complete the characteristic data for each observation angle other than the sampling angle, resulting in a full-angle color characteristic model from 0° to 180°. Specifically, this includes:

[0094] For a specific angle θ i Transformation matrix A(θ) i Using this angle, the two sampling angles θ closest to this angle on both sides. i-1 and θ i+1 Linear interpolation is performed on the matrix; for angle θ i RGB three-channel brightness lookup table LUT(θ) i Also by θ i-1 and θ i+1 The lookup table is used for linear interpolation to obtain the transformation matrix A(θ). i ) and the brightness lookup table LUT(θ) i By combining these, we obtain θ. i Color characterization model at angles; repeat the above process for each angle within the range of 0° to 180° to obtain full-angle color characterization model.

[0095] Step S6: Color Space Conversion. The image to be displayed using a cylindrical raster 3D display device undergoes color space conversion, transforming it to a device-independent standard color space and storing its XYZ information.

[0096] Step S7: Perform color correction from different angles. For example... Figure 6As shown, at each angle, the full-angle color characteristic model established in step S5 is applied, and the XYZ information stored in step S6 is used as the standard to inversely calculate the RGB value of each pixel in the image at each angle, thereby obtaining the corrected three-dimensional display encoding information. The specific process is as follows:

[0097] At a specific angle, the XYZ information of the display image stored in step S6 is combined with the A(θ) matrix corresponding to this angle in the full-angle color characteristic model in step S5 to calculate the single-channel brightness Tr(θ), Tg(θ), and Tb(θ) required for the three RGB channels at this angle. According to the three-channel brightness lookup table LUT(θ) corresponding to this angle stored in step S5, or the "brightness corresponds to RGB driving" curve, the single-channel brightness values ​​Tr(θ), Tg(θ), and Tb(θ) are substituted and interpolated to obtain the driving value RGB(θ) required for this angle. Within the range of 0° to 180°, the above process is repeated for all angles required for 3D encoding to obtain the driving information RGB required for each pixel of the displayed image at all angles.

[0098] Step 8 (S8): Image Reproduction. Using the RGB information of each pixel in the image obtained in Step 7 (S7) at all angles, perform 3D display encoding and display the desired image on a cylindrical raster 3D display device.

[0099] Obviously, those skilled in the art can make various modifications and variations to the embodiments of the present invention without departing from the spirit and scope of the invention. Therefore, if these modifications and variations fall within the scope of the claims of the present invention and their equivalents, the present invention also intends to include these modifications and variations.

Claims

1. A multi-angle color correction method based on cylindrical grating three-dimensional display, characterized in that, Includes the following steps: The first step is to display the test images. Two test images are set: one is a pure white image with RGB values ​​(255, 255, 255), and the other is an RGB image divided into 120 regions, each region being a pure color region with different RGB values. After 3D encoding of the two test images, they are displayed using a cylindrical raster 3D display device. When 3D encoding the RGB images, each pure color region uses the same RGB values ​​at all angles, while the RGB values ​​between different pure color regions are different. The second step is performance testing, which involves spectral measurement of the test images to test the spatial uniformity, temporal stability, and channel additivity of the cylindrical grating 3D display device, and making corrections to the hardware system based on the test results. The third step is to collect data from multiple angles. Using a spectral measurement device, the three-dimensional image displayed in the first step is measured from multiple angles within the range of 0° to 180°. The tristimulus values ​​XYZ can be obtained by integrating the measured spectra. The fourth step is to establish a color characteristic model from different angles. For different spectral measurement angles in the third step, a color characteristic model is established for each angle. At each angle, the RGB values ​​of different pure color areas displayed in the first step are combined with the corresponding XYZ values ​​measured in the third step to establish a mapping relationship from the color characteristics RGB of the device-dependent space to the color characteristics XYZ of the device-independent standard space, which is the color characteristic model from different angles. The fifth step is to establish a full-angle model. The color characteristic model obtained in the fourth step is interpolated to complete the characteristic data of each observation angle other than the sampling angle, so as to obtain a full-angle color characteristic model for full-angle observation within 0° to 180°. The sixth step is color space conversion. The image to be corrected is converted to a device-independent standard color space, and its XYZ information is stored. The seventh step is to perform color correction from different angles. At each angle, the full-angle color characteristic model established in the fifth step is applied. Using the XYZ information stored in the sixth step as the standard, the RGB values ​​of each pixel in the image to be corrected at each angle are calculated in reverse to obtain the corrected three-dimensional display encoding information. The eighth step is image reproduction. The RGB information of the image pixels obtained in the seventh step is used to perform three-dimensional display encoding, and the corrected image is displayed on a cylindrical grating three-dimensional display device.

2. The multi-angle color correction method as described in claim 1, characterized in that, In the first step, the setup of the 120 RGB pure color areas needs to have both single-channel driven and multi-channel driven images, specifically set as follows: Within the single-channel range of 15-255, R, G, and B single-channel driving images are set at intervals of 15; within the single-channel range of 15-255, RG combination, RB combination, and GB combination dual-channel driving images are set at intervals of 15, ensuring that the channel driving values ​​within the combination are equal; within the single-channel range of 0-255, RGB combination three-channel driving images are set at intervals of 15, ensuring that the driving values ​​of the RGB three channels in each combination are equal, resulting in 18 combinations from (0,0,0) to (255,255,255).

3. The multi-angle color correction method as described in claim 2, characterized in that, In the third step, the specific settings for multi-angle spectral measurement are as follows: within the range of 0° to 180° with respect to the 3D display panel, a spectral measurement is performed every 5° interval, that is, a total of 36 measurements are performed for each image.

4. The multi-angle color correction method as described in claim 3, characterized in that, In the fourth step, for each angle, the mapping relationship is established using one of the following methods: three-dimensional lookup table (LUT), matrix transformation, piecewise linear interpolation assuming constant chromaticity (PLCC), piecewise linear interpolation assuming varying chromaticity (PLVC), and gain-bias-gamma model (GOG).

5. The multi-angle color correction method as described in claim 4, characterized in that, In the fifth step, for other angles not at the sampling angle, the color characterization model needs to be interpolated using the characterization models of the sampling angles on both sides.

6. The multi-angle color correction method as described in claim 5, characterized in that, In the fifth step, the interpolation method is either linear interpolation or polynomial interpolation.

7. The multi-angle color correction method as described in claim 5, characterized in that, In the sixth step, the color space conversion of the image to be displayed is accomplished by converting the color information of each pixel, and is achieved through the conversion matrix of different color spaces.

8. The multi-angle color correction method as described in claim 6, characterized in that, In the seventh step, at each angle, the XYZ value of each pixel needs to be converted to RGB value using the color characteristic model corresponding to this angle, so as to obtain the RGB information of each pixel at all angles in the range of 0° to 180°, and use the complete RGB information to perform the three-dimensional display encoding in the eighth step.

9. The multi-angle color correction method according to any one of claims 1-8, characterized in that, The spectral measurement equipment is a spectrometer, a multispectral camera, a spectrophotometer, or a colorimeter.

Images