CN104486603A - Multi-projection color correcting method based on HDR (high dynamic range) imaging - Google Patents

Abstract

The present invention proposes a color correction method based on HDR imaging. A commercial camera is used to perform color correction on the system, and the camera becomes a feedback device in the system by restoring the brightness corresponding curves of each channel of the camera to restore the channels of each projector. The brightness response curve of . Then establish a common color interval, and map the colors of each projector into this interval. At the same time, by adaptively adjusting the target image, the dynamic range of the system can be improved as much as possible, and the picture quality can be improved while ensuring the brightness correction accuracy of the system. After calibration, multiple projectors can work together to project a complete and unified picture.

Description

A multi-projection color correction method based on HDR imaging

technical field

The technology belongs to the technical field of projection mosaic display, and specifically relates to the color correction technology of a multi-projection mosaic system.

Background technique

With the continuous improvement of people's requirements for visual experience and the advancement of engineering projection technology, multi-channel projection technology has been widely used in large-scale events, exhibitions, new product launches and other fields. In a multi-channel projection system, due to the color difference between different projectors and the overlapping of projection areas, etc., the color deviation of the system screen is caused. Eliminating these deviations and allowing multiple projectors to jointly display a complete and unified image is more important. The ultimate goal of the projector system.

The traditional method corrects the color of the system by manually adjusting the color channel of the projector or using expensive optical instruments to analyze the output light of the projector. The former requires a lot of manpower, and the latter is complicated to operate and increases the system cost.

Contents of the invention

The present invention provides a multi-projection color correction method based on HDR imaging, which improves the dynamic range of the system as much as possible through self-adaptive adjustment of the target image, ensures the brightness correction accuracy of the system, and improves the picture quality at the same time.

The present invention is realized through the following technical solutions:

A multi-projection color correction method based on HDR imaging, comprising the following steps:

Step 1. According to different application occasions, the geometric mapping relationship between the camera and the projector pixels is obtained, and a geometric mapping table is established;

Step 2. Let the projector project a red image containing an input value of 0-255, and shoot it through a camera with a fixed aperture, and shoot images under different exposure conditions;

Step 3, smoothing the photographed image described in step 2, taking the central areas of different brightness, and finding the average value, as the camera output value corresponding to this brightness;

Step 4. Use the output values of each camera calculated in step 3, combined with the exposure time, to solve the brightness corresponding curve of the red channel of the camera through an optimization method;

Step 5. Solve the brightness corresponding curves of the green and blue channels of the camera according to the method of steps 2-4;

Step 6. The red channel of the projector outputs a certain red value, and the camera selects an appropriate exposure time to capture the image, gradually increasing the brightness of the projector and capturing the image;

Step 7. For the captured image described in step 6, calculate the average value of the central area of the image, calculate the relative projection brightness through the exposure time, measure the brightness corresponding to all output values, and fit a curve as the brightness response of the projector Curve, use the same method to measure the brightness corresponding curves of other channels of the projector and other projectors;

Step 8. For each channel, select a brightness range that each projector can reach as a common brightness range, and establish a color lookup table for each projector color channel;

Step 9. For the overlapping area between adjacent projectors, calculate the corresponding brightness attenuation weight mask according to the distance between the pixel and the edge;

Step 10. Map the target image to the common color interval, store the geometric mapping table, color lookup table, and brightness attenuation weight mask in the graphics rendering pipeline, and adjust the color of each channel of the projector according to the target image in real time through the image real-time pipeline. Enter the value; the projected image will be an image without geometric and color deviation.

Beneficial effects of the present invention:

The present invention proposes a color correction method based on HDR imaging. A commercial camera is used to perform color correction on the system. By restoring the brightness corresponding curves of each channel of the camera, the camera becomes a feedback device in the system for restoring each projector channel. The brightness response curve of . Then establish a common color interval, and map the colors of each projector into this interval. At the same time, by adaptively adjusting the target image, the dynamic range of the system can be improved as much as possible, and the picture quality can be improved while ensuring the brightness correction accuracy of the system. After calibration, multiple projectors can work together to project a complete and unified picture.

Description of drawings

Fig. 1 is a workflow diagram of the color correction proposed by the present invention.

Detailed ways

The color correction technology in the multi-projection seamless fusion system is mainly responsible for processing the color continuity of the projected picture. For the same color, there should be no perceivable color change when displayed. The color discontinuity that needs to be considered in a multi-projector display system is also divided into the color difference between the inside of a single projector and between projectors. The chromaticity change inside a single projector can usually be ignored, but due to lens vignetting and projection distance The brightness is gradually dimmed from the projection center to the surroundings. The color difference between multiple projectors comes from the inability to accurately control the chromaticity and brightness during the manufacturing process of the projectors, especially for ordinary commercial projectors with relatively low manufacturing costs, there are perceivable color differences among projectors of the same model. Areas of overlap between projectors are significantly brighter than other areas and are the main issue we will address during the color correction process.

The present invention is made up of 4 basic modules in concrete implementation, comprises:

(1) Geometric correction module: For a plane or an approximate plane, the homography matrix or Bezier surface is obtained by projecting the checkerboard to obtain the geometric mapping relationship; for a surface with a relatively regular structure, the Bezier is fitted by surface feature points surface to obtain the geometric mapping relationship; for surfaces with complex and irregular geometry, the geometric mapping relationship is obtained by structured light scanning.

(2) Camera Response Curve Restoration Module: By projecting patterns of different brightness, shooting projection patterns at different exposure times, calculating the relationship between different illuminance values, exposure values and camera output values through optimization methods, and restoring each channel of the camera luminance response curve.

(3) Projector response curve recovery module: use the calibrated camera to shoot projection images corresponding to different input values, and calculate the brightness corresponding curve of each channel of the projector. In order to ensure the measurement accuracy and shorten the measurement time, for each color channel of each projector, a total of 16 images are projected, each image contains four luminances, the area close to the center of the projection is taken, and the area average is obtained after smoothing value to calculate the relative brightness corresponding to that input value.

(4) Real-time mapping module for each color channel of the projector: After obtaining the brightness response curve of the projector, for each color channel, a common color range is obtained to ensure that each projector can output the brightness within the common range . After the target image is adaptively adjusted, the target color value is used to calculate the input brightness in real time through a lookup table stored in the graphics rendering pipeline.

The multi-channel projection color correction technology based on HDR imaging includes the following steps:

1. Obtain the geometric mapping relationship between the camera and projector pixels according to different application occasions.

2. Let the red channel of the projector project images with brightness from 0-255. Captured by a video camera with a fixed aperture, five images were captured each under different exposure conditions.

3. Perform smoothing processing on the captured image. For the image under the same exposure time, take the center brightness of the input same brightness part.

4. Combined with the exposure time, use the following equation to solve the brightness corresponding curve of the red channel of the camera through the optimization method, where Z represents the camera output value, E represents the illuminance value actually received by the pixel unit, and t represents the exposure time.

g(Z) ₌ ln(E)+lnt

5. Solve the brightness corresponding curves of the green and blue channels of the camera according to the method of steps 2-4.

6. The red channel of the projector outputs a certain red value, and the camera selects an appropriate exposure time to capture this image, gradually increases the brightness of the projector and captures images, and captures 5 images for each input value.

7. Take the value of the center area of the captured image as the value, and calculate the relative projection brightness through the exposure time and the camera output value. After measuring the luminance corresponding to all output values, a curve is fitted as the projector luminance response curve. Use the same method to measure the brightness corresponding curves of other channels of the projector and other projectors.

8. For each channel, select a brightness range that each projector can achieve as a common brightness range, and build a lookup table for each projector color channel.

9. For overlapping regions, calculate the corresponding attenuation weight mask according to the distance of the pixel from the edge. Since the brightness response curve of the projector has been obtained, the attenuation weight here is calculated using a linear attenuation method.

10. Map the target image into a common color interval. Store geometry maps, color lookup tables, and edge attenuation masks in the graphics rendering pipeline. Through the designed image real-time pipeline, the input value of each channel of the projector is adjusted in real time according to the target image.

Although the present invention is described with reference to the preferred embodiments, the above examples do not constitute a limitation of the protection scope of the present invention, and any modifications, equivalent replacements and improvements within the spirit and principles of the present invention should be included in the scope of the present invention. within the scope of the claims.

Claims (3)

1. A multi-projection color correction method based on HDR imaging, characterized in that, comprising the following steps: Step 1. According to different application occasions, the geometric mapping relationship between the camera and the projector pixels is obtained, and a geometric mapping table is established; Step 2. Let the projector project a red image containing an input value of 0-255, and shoot it through a camera with a fixed aperture, and shoot images under different exposure conditions; Step 3, smoothing the photographed image described in step 2, taking the central areas of different brightness, and finding the average value, as the camera output value corresponding to this brightness; Step 4. Use the output values of each camera calculated in step 3, combined with the exposure time, to solve the brightness corresponding curve of the red channel of the camera through an optimization method; Step 5. Solve the brightness corresponding curves of the green and blue channels of the camera according to the method of steps 2-4; Step 6. The red channel of the projector outputs a certain red value, and the camera selects an appropriate exposure time to capture the image, gradually increasing the brightness of the projector and capturing the image; Step 7. For the captured image described in step 6, calculate the average value of the central area of the image, calculate the relative projection brightness through the exposure time, measure the brightness corresponding to all output values, and fit a curve as the brightness response of the projector Curve, use the same method to measure the brightness corresponding curves of other channels of the projector and other projectors; Step 8. For each channel, select a brightness range that each projector can reach as a common brightness range, and establish a color lookup table for each projector color channel; Step 9. For the overlapping area between adjacent projectors, calculate the corresponding brightness attenuation weight mask according to the distance between the pixel and the edge; Step 10. Map the target image to the common color interval, store the geometric mapping table, color lookup table, and brightness attenuation weight mask in the graphics rendering pipeline, and adjust the color of each channel of the projector according to the target image in real time through the image real-time pipeline. Enter the value; the projected image will be an image without geometric and color deviation. 2. A multi-projection color correction method based on HDR imaging as claimed in claim 1, wherein the solution of the camera output value in step 4 adopts the following method: wherein Z represents the camera output value, and E represents the pixel unit actually received To the illuminance value, t represents the exposure time g(Z)=ln(E)+lnt. 3. A multi-projection color correction method based on HDR imaging according to claim 1 or 2, wherein 5 images are captured each time an image is captured.