KR20080105272A - Color correction method for the projector - Google Patents

Abstract

The present invention relates to a color correction method for a projector, and more specifically, (1) obtaining a color fitting parameter for each pixel of a screen on which an image is projected from a projector, and (2) Spatially smoothing each of the color adjustment parameters using a smoothing filter, (3) linearizing the RGB values of the input image through a device characterization process of the projector, and (4) linearizing RGB Converting values (R ₀ , G ₀ , B ₀ ) to values of device independent CIELUV color space (L ₀ , u ₀ , v ₀ ), and (5) using each smoothed color adjustment parameter Mapping the (u ₀ , v ₀ ) values to the gamut that each pixel can represent to obtain the mapped (u ₁ , v ₁ ) values, and (6) performing luminance fitting. The luminance value L adjusted from the luminance value L ₀ by ₁ ) obtaining, and (7) obtaining a corrected RGB value (R ₁ , G ₁ , B ₁ ) of the input image from the obtained values of the corrected CIELUV color space (L ₁ , u ₁ , v ₁ ). It characterized by including the configuration.

According to the projector color correction method of the present invention, each color adjustment parameter is obtained for each pixel of the screen on which the image is projected from the projector and spatially smoothed, and then the color gamut is mapped using the obtained smoothed color adjustment parameters. By correcting the input image through the process, it is possible to obtain an image of improved quality by extending the color gamut of the image to be reproduced while maintaining the advantages of the existing color correction method.

Description

Color correction method for projectors {A CHROMINANCE CORRECTION METHOD FOR A PROJECTOR}

1 is a view for explaining the necessity of gamut mapping due to color gamut mismatch.

FIG. 2 is a diagram showing a result of projecting a projector gamut in a CIELUV color space corresponding to one pixel to a plane having a luminance value of 0; FIG.

3 is a view for explaining a problem with the Ashdown method, that is, the color gamut that can be expressed is greatly reduced.

4 shows a flowchart of a color correction method for a projector according to an embodiment of the present invention for solving the problem of the Ashdown method.

5 shows a comparison of the color adjustment parameters s for a column of sample images with respect to the Ashdown method, the proposed method (before smoothing), and the proposed method (after smoothing).

6 shows a screen used in a projector color calibration experiment.

7 is a view showing the results of measuring the eight colors projected on the screen with a camera.

8 is a view showing the color gamut corresponding to one column of the screen used in the experiment.

9A is a diagram illustrating a test pattern used in an experiment, FIG. 9B is a diagram illustrating an output image when there is no correction, and FIG. 9C is a diagram illustrating an input image after applying the correction method of the present invention. 9E is a diagram showing an output image by an existing method (Ashdown method) and an output image by the method of the present invention, respectively.

FIG. 10 is a graph showing the results of the experiment in FIG. 9, and FIGS. 10A to 10F are u values of the output image before correction, v values of the output image before correction, and output image after correction by the Ashdown method, respectively. U value, v value of the output image after correction by the Ashdown method, u value of the output image after correction by the method of the present invention, v value of the output image after correction by the method of the present invention.

11 is a view showing a natural image used in the experiment.

12 to 14 are views showing the results of applying the method proposed in the present invention to the three natural images of FIG. 11 compared with the results of applying the conventional Ashdown method.

FIG. 15 is a diagram showing color adjustment parameters s for experimental images 1, 2, and 3 compared with the conventional Ashdown method. FIG.

<Explanation of symbols for main parts of the drawings>

S10: obtaining color adjustment parameters for each pixel of the screen

S20: Spatially Smoothing Color Adjustment Parameters

S30: Linearizing the RGB values of the input image through the device characterization process of the projector

S40: converting linearized RGB values to values of device independent CIELUV color space

S50: color gamut mapping the color values (u, v) of the CIELUV color space using the smoothed color adjustment parameters

S60: performing luminance adjustment

S70: obtaining a corrected RGB value of the input image from the value of the corrected CIELUV color space

The present invention relates to a color correction method for a projector, and more specifically, to obtain each color adjustment parameter (chrominance fitting parameter) for each pixel of the screen on which the image is projected from the projector, and to smooth it spatially, the obtained smoothed By correcting the input image by mapping the color gamut using each color adjustment parameter, it is possible to obtain an image of improved quality by extending the gamut of the image to be reproduced while maintaining the advantages of the existing color correction method. A color correction method for a projector.

As the size of the projector is smaller and the price is lowered, research on the field of image display using the projector is being actively conducted. As the use of projectors increases, the projection of images on a white screen in the dark room, which is the ideal environment, can be projected to various environments, such as walls or paper, not on a normal screen, or when there are various colors on the screen, or bent like a curtain. Projections on such surfaces often occur. Unlike the original image, the projected images cause geometric distortion, deterioration of brightness and color quality, and thus, a technology development for reproducing a natural image so that the projected image is closer to the original image is required.

The method of correcting the color distortion due to the characteristics of the screen is one of the active research areas. When the color is present on the screen, a phenomenon in which the reproduced colors appear differently depending on the color occurs, because a difference in the range that can express the color, that is, the gamut of each pixel of the screen. Therefore, various gamut mapping methods have been studied for color reproduction close to the original image. Color gamut mapping is a method of mapping color gamuts between different devices, and an object thereof is to reproduce an output image having a color characteristic similar to that of an original image. In order to map the color gamut, it is necessary to first determine which color space to use. Commonly used color spaces include CIEXYZ, CIELAB, and CIELUV. As the gamut mapping method, there are a color clipping method and a color compression method according to the mapping type, and it may be divided into four methods according to the mapping direction. The gamut cutting method refers to a method of mapping a color outside the reproduction range to a narrow gamut boundary, and the gamut compression method refers to a method of changing all colors in the gamut. In the gamut compression method, a linear function or a nonlinear function may be used. Recently, a correction method using a spatially uniform chrominance transformation and a spatially varying luminance transformation depending on the input image has been proposed. This method has the advantage of adaptive color gamut mapping depending on the input image and the reduction of non-uniform brightness difference of the projected image effectively by applying spatially varying brightness conversion. There is a problem that the range of colors that can be represented is greatly reduced because it is a method of mapping to a common area of color gamuts of a pixel. Therefore, there is a need to develop a method that can solve the problem of reducing the range of reproducible colors while taking advantage of the spatially uniform color conversion and spatially varying luminance conversion depending on the input image.

The present invention is derived from the above problem and need recognition, and after obtaining each color adjustment parameter for each pixel of the screen on which the image is projected from the projector, and smoothing it spatially, using each obtained smoothed color adjustment parameter By correcting the input image through the process of mapping the gamut, it provides a color correction method for a projector that can obtain an image of improved quality by extending the gamut of the image to be reproduced while maintaining the advantages of the existing color correction method. The purpose.

According to a feature of the present invention for achieving the above object, a color correction method for a projector,

(1) obtaining a color fitting parameter for each pixel of the screen on which the image is projected from the projector;

(2) spatially smoothing each of the obtained color adjustment parameters using a smoothing filter;

(3) linearizing the RGB values of the input image through a device characterization process of the projector;

(4) converting the linearized RGB values (R ₀ , G ₀ , B ₀ ) into values (L ₀ , u ₀ , v ₀ ) of the device independent CIELUV color space;

(5) Using the smoothed respective color adjustment parameters, map the (u ₀ , v ₀ ) values to the gamut that each pixel can represent to obtain the mapped (u ₁ , v ₁ ) values. step;

(6) performing luminance fitting to obtain the adjusted luminance value L ₁ from the luminance value L ₀ ; And

(7) obtaining a corrected RGB value (R ₁ , G ₁ , B ₁ ) of the input image from the obtained values of the corrected CIELUV color space (L ₁ , u ₁ , v ₁ ). It features.

Preferably, in step (2), a Gaussian mask, which can be expressed as the following equation, is used as the smoothing filter, but the constant k value is advantageously set to 0.93.

Here, i and j represent an image position.

Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings.

First, the reason why the gamut mapping is required, the existing gamut mapping method and its problems will be briefly described.

By performing a device characterization process of the projector and a device independent color conversion process of input image values, a gamut mapping for applying a color of an input image to a color of an output image projected by the projector is applied. In this case, when the color gamut of the original device and the color gamut of the device to be reproduced are inconsistent, a gamut mapping process for colors deviating from the gamut of the device to be reproduced is required. 1 is a view for explaining the necessity of gamut mapping due to color gamut mismatch. In FIG. 1, a region (a region including red dots) which is not included in the color gamut (triangle indicated by a solid line) of the device to be reproduced (the triangle including red dots) of the original device's gamut (triangle indicated by dotted lines) is the gamut of the device to be reproduced through the gamut mapping process It must be thought of as myself. Arrows in FIG. 1 indicate such color gamut mapping.

As mentioned above, the purpose of gamut mapping is to solve the deviation of the size, shape, and position existing between the original and the gamut of the device to be reproduced through an appropriate mapping method. Therefore, various thought methods have been studied according to thought elements, and they can be classified into methods by thought type and thought direction. It can also be divided into image independent gamut mapping independent of the input image and content dependent gamut mapping dependent on the input image. Image-independent gamut mapping is a method that has been studied a lot. It means mapping between the gamut of the original device and the gamut of the device to be reproduced. Since only the mapping between the color gamuts of the two devices is considered, human visual adaptation cannot be considered depending on the color of the input image. In order to solve such drawbacks, research on image-dependent mapping methods has recently begun. The image dependent mapping method refers to mapping between the gamut of a given input image and the gamut of the device to be reproduced, and an adaptive gamut mapping depending on the input image is possible.

Among the image dependent gamut mapping methods, a correction method using spatially uniform color conversion and spatially varying luminance conversion depending on the input image has been proposed by Ashdown et al. This method (hereinafter referred to as the Ashdown method) converts the values of the input image into values of the device independent CIELUV color space and then starts gamut mapping. Unlike the existing methods, the upper and lower limits of the luminance values of the u and v values to be reproduced are defined, and smooth brightness changes in the spatial domain can be expected.

The first step of the Ashdown method is to linearize the RGB values of the input image and convert them to (L ₀ , u ₀ , v ₀ ), which are values of the device-independent CIELUV color space. Here, linearizing the RGB values of the input image means modifying the RGB values of the input image through a device characterization process of the projector.

In the next step, two-dimensional projection of the gamut that can be reproduced for each pixel of the projector is carried out in order to perform chrominance fitting. FIG. 2 is a diagram illustrating a result of projecting a projector gamut in a CIELUV color space corresponding to one pixel to a plane having a luminance value of zero. The boundary of the gamut in each pixel is defined by six lines connecting red, yellow, green, cyan, blue, and magenta, which are indicated by triangles in FIG. 2, and the white triangle inside the gamut represents a white point. The relationship between the chrominance to be reproduced and the original chrominance may be defined by a spatially uniform linear transformation equation as shown in Equation 1 below.

In the Ashdown method, we define an objective function that can reduce the error between the original color and the mapped gamut and at the same time widen the gamut of the mapped gamut and apply a nonlinear optimization method. In the present invention, a method of obtaining a variable in each pixel and smoothing the same is applied to prevent a sudden change due to screen color change while widening the gamut.

As mentioned above, the Ashdown method has a big disadvantage in the color correction method. That is, since the color of the input image is mapped to a common area of the color gamuts of each pixel of the screen, the range of colors that can be expressed is greatly reduced. 3 is a diagram for explaining a problem of the Ashdown method, that is, the color gamut that can be expressed is greatly reduced. In FIG. 3, a triangle indicated by a dotted line indicates a color distribution of a specific input image, and two triangles indicated by a solid line indicate a color gamut of two different pixels in which colors of a screen are reflected. Since the gamut of the original image is outside the gamut of the projector, a gamut mapping process is required. When applying the Ashdown method, the gamut of the original image is a common area of gamuts that can be represented by each pixel on the screen (two different pixels in FIG. It is mapped to (indicated by diagonal lines). That is, although the colors of the area indicated by the horizontal line and the area indicated by the vertical line can be expressed, the color that can be reproduced is greatly reduced by mapping to the common color gamut (the area indicated by the diagonal lines) of all the pixels on the screen.

4 is a flowchart illustrating a color correction method for a projector according to an embodiment of the present invention for solving the problem of the Ashdown method. As shown in FIG. 4, in the color correction method according to an exemplary embodiment of the present invention, obtaining color adjustment parameters for each pixel of the screen (S10), and spatially smoothing the color adjustment parameters (S20). Linearizing the RGB values of the input image through the device characterization process of the projector (S30), converting the linearized RGB values into values of the device-independent CIELUV color space (S40), and smoothing color adjustment parameters in step S20 Color gamut mapping the color values u and v of the CIELUV color space (S50), performing the brightness adjustment (S60), and corrected RGB values of the input image from the values of the corrected CIELUV color space using Obtaining step (S70). Looking at the state of the signal in each step, as shown to the right of each step in FIG. 4, in step S30 (R ₀ , G ₀ , B ₀ ), in step S40 (L ₀ , u ₀ , v _0), and the step S50 in the (L _0, u _1, v _1), in the step _{S60 (L 1, u 1,} v 1), in step _{S70 (R 1, G 1,} B 1). In the following, each step will be described in more detail.

First, with regard to step S30, the device characterization technique of the improved projector proposed in the present invention will be described. That is, a device characterization technique for reducing the white enhancement phenomenon of a digital lighting processing (DLP) projector by considering interference between RGB channels will be described. First, projectors project uniform images including eight images of saturated red, green, blue, cyan, magenta, yellow, black, and white. Debevec et al.'S method proposed in 1997 takes several pictures with different exposure times of images projected by a digital camera, and then obtains the corresponding brightness value at each pixel of the screen. The next step is to measure the projector's device-independent XYZ values using a colorimeter or chromameter, a commonly used color measurement instrument. Using the measured value, the conversion matrix M between the brightness value at each pixel, that is, the RGB value measured by the camera and the XYZ value measured by the colorimeter, can be obtained through Equation 2 below.

Here, M is a 3x3 transformation matrix and has a different value for each pixel of the screen.

Next, the color correction method considering the spatial characteristics of the input image and the screen will be described in earnest. In the Ashdown method, since the linear conversion equation is applied to all pixels of the input image in the uv color space of CIELUV, the color adjustment parameters have the same value for each pixel of the screen. That is, the Ashdown method compresses the original color gamut into a common color gamut of each pixel of the projector to be reproduced. Since only the common gamut portion is used, the range of colors that can be expressed is greatly reduced. Therefore, in the present invention, in order to further improve the color of the image to be reproduced, a method of expressing the color that can be represented in each pixel to the maximum and gradually changing it according to the position is proposed. In the method proposed in the present invention, (u ₀ , v ₀ ) of the input image is mapped to the color gamut represented by each pixel, and the color gamut of the color to be reproduced can be extended by applying different color adjustment parameters to each pixel. Will be. However, (u ₁ , v ₁ ) obtained by the above method has a problem in that it can be changed rapidly spatially because s, a, b are obtained independently in the uv color space for each color gamut. Therefore, in the present invention, in order to solve this problem, the next step is to gently change the color adjustment parameters in the spatial domain so that both the screen and the input image characteristics can be considered.

FIG. 5 is a diagram illustrating color adjustment parameters s for one column of a sample image in comparison with the Ashdown method, the proposed method (before smoothing), and the proposed method (after smoothing). As can be seen in FIG. 5, the color adjustment parameter ('global s' in FIG. 5) obtained in the Ashdown method significantly reduces the (u, v) values of the original image, whereas the color adjustment obtained in the method proposed in the present invention. The parameter 'each s' of FIG. 5 has a different value for each pixel of the screen, and the smoothed color adjustment parameter ('after smoothing' of FIG. 5) that is spatially smoothed is larger than the color adjustment parameter obtained in the Ashdown method. Value, resulting in a wider gamut representation. As a smoothing filter, a Gaussian mask, such as Equation 3, may be used.

Here, i and j represent an image position, and the k value is set to 0.93.

Next, after luminance fitting (S60), the corrected RGB values (R ₁ , G ₁ , B ₁ ) of the input image are calculated from the finally obtained (L ₁ , u ₁ , v ₁ ). (S70). Step S70 may be implemented by the reverse process of step S40.

The actual experimental results of the color correction method for the projector described above will be described below. First, the process of measuring the gamut of the projector used in the experiment will be described.The results of experiments with primary and secondary colors, ie, red, green, blue, cyan, magenta, and yellow, will be compared and analyzed. Let's take a look at the results applied to the images. 6 is a diagram illustrating a screen used in a projector color calibration experiment.

(1) color gamut measurement experiment

Eight images of red, green, blue, cyan, magenta, yellow, black and white were used to measure the color gamut for each pixel on the screen. 7 is a diagram illustrating a result of measuring eight colors projected on a screen by a camera.

Next, measure the XYZ values for 8 colors red, green, blue, cyan, magenta, yellow, black, and white using a spectroradiometer, and then use the method proposed by Debevec et al. And calculate the XYZ value at each pixel of the screen as shown in Equation 4 below.

In Equation 4, M _white denotes a conversion of a relationship between an XYZ value measured by a spectroradiometer and a brightness value on a white screen, and ij denotes a position of each pixel of the screen. Finally, XYZ values corresponding to the input image RGBCMY are obtained from Equation 4, and then, using these values, a color gamut considering the characteristics of the screen is obtained for each pixel. 8 is a diagram illustrating color gamuts corresponding to one column of a screen used in an experiment. In FIG. 8, the small circle is the white point of each color gamut, and since the color of the screen used is three, it can be seen that it is distributed in three areas.

(2) projector color calibration experiment for test pattern

Next, the result of the color correction experiment on the test pattern will be described. FIG. 9A is a diagram illustrating a test pattern used in an experiment, FIG. 9B is a diagram illustrating an output image when there is no correction, and FIG. 9C is a diagram illustrating an input image after applying the correction method of the present invention. 9E is a diagram showing an output image by an existing method (Ashdown method) and an output image by the method of the present invention, respectively. Referring to FIG. 9, first, it can be seen from FIG. 9B that brightness and color characteristics of the projector are reflected. In other words, the left and right parts with the green and pink colors on the screen are relatively dark, and the parts projected on the white screen are relatively bright. Also, in the green and pink screen area, there are many color differences with the original image. I can confirm that In addition, the white screen part also shows a color difference from the original image due to the influence of the projector's gamut, and the phenomenon of darkening toward the edge of the screen due to the vignetting phenomenon of the projector can be identified. By comparing FIG. 9D with FIG. 9E, it can be clearly confirmed that the result image by the method proposed in the present invention is closer to the original image than the result image by the conventional Ashdown method.

FIG. 10 is a graph showing the results of the experiment in FIG. 9, and FIGS. 10A to 10F are u values of the output image before correction, v values of the output image before correction, and output image after correction by the Ashdown method, respectively. U value, the v value of the output image after correction by the Ashdown method, the u value of the output image after correction by the method of the present invention, and the v value of the output image after correction by the method of the present invention. For each of the six colors used in the test image, the area was divided into pink, white and light green screens to obtain the average RGB values for the six colors. The average RGB values were converted back to (u ₀ , v ₀ ). (U, v) values were obtained for the output image before correction and the output image after correction, respectively. 10, it can be seen that the method proposed by the present invention shows an improved result than the conventional Ashdown method and is significantly closer to the original image than before the correction. Table 1 below shows the comparison of the u and v coordinate values of the original image, the output image before correction, and the output image after correction by the method according to the present invention for the six input colors.

u Original video Before calibration After calibration Lime green screen White screen Pink screen Lime green screen White screen Pink screen red 0.9201 0.4892 0.4958 0.9126 0.6052 0.9315 0.7494 magenta 0.6017 0.2519 0.1607 0.2938 0.4550 0.4094 0.5192 blue -0.0836 -0.1897 -0.2176 -0.0365 -0.1511 -0.0541 -0.1027 cyan -0.1839 -0.4754 -0.4196 -0.2504 -0.3887 -0.4591 -0.3431 green -0.5572 -0.4388 -0.4479 -0.2344 -0.5240 -0.5235 -0.3862 yellow 0.0372 -0.1613 -0.1129 0.1675 -0.1310 -0.0629 -0.0354

v Original video Before calibration After calibration Lime green screen White screen Pink screen Lime green screen White screen Pink screen red 0.2041 0.2246 0.0157 0.2208 -0.0102 0.1839 0.0540 magenta -0.7438 -0.3050 -0.9328 -0.6216 -0.6447 -0.9028 -0.7543 blue -1.0126 -0.4304 -0.9414 -0.8160 -0.8843 -1.0992 -0.9776 cyan -0.1839 -0.3632 -0.1317 -0.0138 -0.2205 -0.2877 -0.2261 green 0.7440 0.7183 0.4267 0.5481 0.3781 0.4206 0.2677 yellow 0.7839 0.6995 0.4571 0.5648 0.6636 0.5876 0.6634

Using the values shown in Table 1, by the following equation 5, when the correction by the conventional method, before the correction, the error when the correction by the method proposed in the present invention was calculated. As a result of the calculation, it was confirmed that the error can be reduced to 50% or less when using the method proposed by the present invention compared to the conventional method.

Here, n represents the number of pixels of the input image.

(3) Projector color correction experiment applied to real image

Finally, in order to confirm a more substantial improvement effect, the method proposed in the present invention is applied to the natural image and the experimental results are analyzed while comparing with the existing method. The image to be used for the experiment is the same as that of FIG. 11, and its size is 200 × 150.

12 to 14 are diagrams showing the experimental results of applying the method proposed in the present invention to the three natural images of FIG. 11 in comparison with the results of applying the conventional Ashdown method. 12 to 14A are diagrams illustrating the original image shown in FIG. 11 again. 12 to 14 (b) show an image in which the original image is projected as it is, and it can be clearly seen that the white screen area becomes bright and the light green and pink screen areas become dark. You can also see the difference in color, which is more noticeable at the border of the screen color. 12 to 14 (c) show input images corrected by applying the algorithm proposed in the present invention, and show different input brightness and colors in the green-green, white and pink screen regions, respectively. In other words, it is relatively dark in the white screen area and bright in the colored screen area. If the corrected input image of FIG. 12 to FIG. 14C is projected to the projector, the corrected output image may be obtained as shown in FIG. 12 to FIG. For performance comparison, FIGS. 12 to 14 (d) show the results of applying the conventional Ashdown method. As a result of applying the existing Ashdown method, the color difference is slightly reduced, but the color difference is partially unnatural. In addition, there is still a difference in brightness between the colored screen and the white screen, resulting in an overall dark output image. On the contrary, when looking at the image to which the method proposed in the present invention is applied, the overall brightness becomes smooth and smooth brightness change can be confirmed. In addition, because the colors that can be reproduced are used as much as possible, the image quality of the output color is improved to be close to the original image.

In particular, in FIG. 12, although the colors of the red flowers are slightly different for each screen area in the existing Ashdown method, the results of applying the method proposed in the present invention can be seen to be closer to the colors of the original image as a whole. For existing buildings, the degree of improvement around the screen color boundaries is more evident. In FIG. 13, it can be seen that the method proposed by the present invention exhibits superior performance than the conventional Ashdown method. In this case, the difference in performance can be easily seen by comparing the leaf area of flowers in the green and pink screen area. In the last image of FIG. 14, the brightness of the original image itself tends to become dark from left to right. As a result of projecting the original image as it is, it can be seen that the part that should be darkened became bright due to the effect of the white screen. The result of applying the existing Ashdown method also does not properly reflect the tendency of the brightness of the original image. However, looking at the results of applying the method proposed in the present invention, it can be seen that this unnaturalness is greatly improved.

As described above, the proposed method shows improved performance over the conventional Ashdown method by considering both spatial characteristics of the original image and the screen.

FIG. 15 is a diagram illustrating color adjustment parameters s of experimental images 1, 2, and 3 compared with the conventional Ashdown method. First, the global s parameter using the existing Ashdown method is indicated by dashed lines, showing the smallest value in all three images. This reflects the characteristics of the existing Ashdown method, which maps to a common color that all pixels can represent. The result of the adjustment for each pixel by applying the method proposed in the present invention is shown by a dotted line, as can be seen in the drawing, it is difficult to expect a smooth change between adjacent pixels. Therefore, as the solid filter passes through the spatial domain, the characteristics in the spatial domain are well reflected, and a larger value than the parameter by the conventional Ashdown method can be expected.

The present invention described above may be variously modified or applied by those skilled in the art, and the scope of the technical idea according to the present invention should be defined by the following claims.

According to the projector color correction method of the present invention, each color adjustment parameter is obtained for each pixel of the screen on which the image is projected from the projector and spatially smoothed, and then the color gamut is mapped using the obtained smoothed color adjustment parameters. By correcting the input image through the process, it is possible to obtain an image of improved quality by extending the color gamut of the image to be reproduced while maintaining the advantages of the existing color correction method.

Claims (2)

(1) obtaining a color fitting parameter for each pixel of the screen on which the image is projected from the projector; (2) spatially smoothing each of the obtained color adjustment parameters using a smoothing filter; (3) linearizing the RGB values of the input image through a device characterization process of the projector; (4) converting the linearized RGB values (R ₀ , G ₀ , B ₀ ) into values (L ₀ , u ₀ , v ₀ ) of the device independent CIELUV color space; (5) Using the smoothed respective color adjustment parameters, map the (u ₀ , v ₀ ) values to the gamut that each pixel can represent to obtain the mapped (u ₁ , v ₁ ) values. step; (6) performing luminance fitting to obtain the adjusted luminance value L ₁ from the luminance value L ₀ ; And (7) obtaining a corrected RGB value (R ₁ , G ₁ , B ₁ ) of the input image from the obtained values of the corrected CIELUV color space (L ₁ , u ₁ , v ₁ ) Color correction method for a projector comprising a. The method of claim 1, In the step (2), using a Gaussian mask that can be expressed as the following equation with the smoothing filter, the constant k value is set to 0.93.

Here, i and j represent an image position.

Images