WO2022170482A1 - 一种颜色标定方法、控制器、相机、电子设备及存储介质 - Google Patents

一种颜色标定方法、控制器、相机、电子设备及存储介质 Download PDF

Info

Publication number
WO2022170482A1
WO2022170482A1 PCT/CN2021/076275 CN2021076275W WO2022170482A1 WO 2022170482 A1 WO2022170482 A1 WO 2022170482A1 CN 2021076275 W CN2021076275 W CN 2021076275W WO 2022170482 A1 WO2022170482 A1 WO 2022170482A1
Authority
WO
WIPO (PCT)
Prior art keywords
color
image
gradient
controller
shape
Prior art date
Application number
PCT/CN2021/076275
Other languages
English (en)
French (fr)
Inventor
曾志豪
何健
罗如君
Original Assignee
深圳市大疆创新科技有限公司
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 深圳市大疆创新科技有限公司 filed Critical 深圳市大疆创新科技有限公司
Priority to PCT/CN2021/076275 priority Critical patent/WO2022170482A1/zh
Publication of WO2022170482A1 publication Critical patent/WO2022170482A1/zh

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis

Definitions

  • the present application relates to the technical field of image processing, and in particular, to a color calibration method, a controller, a camera, an electronic device and a storage medium.
  • the camera records the color of the photographed object through the image sensor, which is the light-sensitive element. Due to the different spectral responses of different image sensors and different processing algorithms for photoelectric signals, the color of the photographed object in the camera is different from the color seen by the human eye. In order to restore the color perceived by the human eye, the camera often needs to be color calibrated before leaving the factory, so that the color of the object in the captured image is consistent with its real color.
  • one of the objectives of the present application is to provide a color calibration method, controller, camera, electronic device and storage medium, which can improve the accuracy of the color correction model, that is, the generalization ability, thereby improving the color restoration accuracy.
  • a color calibration method comprising:
  • a color correction model is generated based on the color samples and the color reference values to color calibrate the image sensor.
  • a controller comprising:
  • memory for storing processor-executable instructions
  • the processor is configured to:
  • a color correction model is generated based on the color samples and the color reference values to color calibrate the image sensor.
  • a third aspect provides a camera including an image sensor and the controller of the second aspect.
  • an electronic device in a fourth aspect, includes the controller according to the second aspect, and the electronic device is configured to perform color calibration on a camera.
  • a computer storage medium on which computer instructions are stored, and when the instructions are executed by a processor, any one of the above methods is implemented.
  • the present application provides a color calibration method, a controller, a camera, an electronic device and a storage medium. After photographing a gradient color card and performing shape correction processing on the photographed image, at least one pixel point with different pixel values is extracted from the image as Color sample set; since color samples representing a large number of colors in the color space can be extracted from the gradient color card, the color correction model generated based on the color samples and the reference values corresponding to the color samples has high accuracy, thereby improving the color restoration accuracy.
  • Figure 1 is a 24-color standard color card and a 140-color ColorChecker SG color card in the related art.
  • FIG. 2 shows a color calibration method according to an exemplary embodiment of the present application.
  • FIG. 3 is a TE234 gradient color card shown in the present application according to another exemplary embodiment.
  • FIG. 4 is a color calibration method shown in the present application according to another exemplary embodiment.
  • Fig. 5 shows a color calibration method according to another exemplary embodiment of the present application.
  • FIG. 6 is a controller according to another exemplary embodiment of the present application.
  • FIG. 7 is a camera according to another exemplary embodiment of the present application.
  • FIG. 8 is an electronic device according to another exemplary embodiment of the present application.
  • At least one means one or more, and “plurality” means two or more.
  • And/or which describes the association relationship of the associated objects, means that there can be three kinds of relationships, for example, A and/or B, it can mean that A exists alone, A and B exist at the same time, and B exists alone, where A, B can be singular or plural.
  • the character “/” generally indicates that the associated objects are an “or” relationship.
  • At least one item(s) below” or similar expressions thereof refer to any combination of these items, including any combination of single item(s) or plural items(s).
  • At least one (a) of a, b, or c may represent: a, b, c, a and b, a and c, b and c, or a and b and c, where a, b, c Can be single or multiple.
  • the camera Before the camera leaves the factory or before the user uses the camera, it is often necessary to perform color calibration with a parameter adjustment tool, so that the color of the object in the image captured by the camera is consistent with the real color.
  • the camera shoots a standard color card such as a 24-color standard color card (left in Figure 1) or a 140-color ColorChecker SG (right in Figure 1) under a standard light source.
  • the standard color card includes multiple color blocks, and the color standard value of each color block under the standard light source is a known empirical value. Therefore, by combining the camera color output value of each color block and the color standard value corresponding to the color block As a training sample, the color calibration model can be trained to complete the calibration of the camera color.
  • the color of the object in the output image can be consistent with the real color.
  • the number of commonly used standard color cards is limited.
  • the 24-color standard color card has only 24 color blocks, that is, the number of training samples that can be provided is only 24, and the number of training samples that the 140-color ColorChecker SG can provide is only 140.
  • the number of training samples available from commonly used standard color cards limited, but these training samples only represent a small number of colors in the color space.
  • the accuracy of the trained color calibration model is not high, and it is easy to overfit, resulting in weak generalization ability of the model.
  • the present application provides a color calibration method, which can be applied to cameras, electronic devices connected to cameras, and other electronic devices equipped with image sensors.
  • the camera includes an image sensor, that is, a photosensitive element, which is used to record the color of the photographed object, and the electronic device connected to the camera can read the data or parameters stored by the camera, and can modify the camera parameters or write data, such as the camera parameter tuning tool.
  • the color calibration method includes:
  • Step 210 acquiring the image of the gradient color card captured by the image sensor
  • Step 220 Perform shape correction processing on the image to obtain a corrected image, so that the shape of the corrected image matches the shape of the gradient color card;
  • Step 230 Extracting at least one pixel point with different pixel values from the corrected image as a color sample set, and acquiring the color reference value corresponding to each color sample in the color sample set;
  • Step 240 Generate a color correction model based on the color sample and the color reference value to perform color calibration on the image sensor.
  • a color calibration method In a color calibration method provided by the present application, after taking a gradient color card and performing shape correction processing on the captured image, at least one pixel point with different pixel values is extracted from the image as a color sample set; Color samples representing a large number of colors in the color space can be extracted, so the color correction model generated based on the color samples and the reference values corresponding to the color samples has high accuracy, thereby improving the color restoration accuracy.
  • the image sensor can shoot gradient color cards under standard light sources such as D65, D50, and TL84, and the color standard values of gradient color cards under standard light sources are known.
  • any other light source of a gradient color card with known color standard values can also be used for shooting, which is not limited in this application.
  • the gradient color card may be a TE234 gradient color card as shown in FIG. 3, and of course, it may be a commercial or custom gradient color card with known color standard values, and is not limited to a color gradient color card, It can also be a grayscale color card, and the shape of the gradient color card can also be various, which is not limited in this application.
  • the following description will take the gradient color card TE234 shown in Figure 3 as an example.
  • the shooting angle will cause the captured image to be distorted.
  • the shape of the gradient color card may be deformed in the captured image.
  • the gradient color card is a rectangle, but in the captured image, it becomes a parallelogram instead of a standard rectangle. .
  • the shape of the image has an important influence on the color reference value corresponding to the color sample obtained in the subsequent step 230, so it is necessary to perform shape correction processing on the image first, so that the shape of the corrected image matches the shape of the gradient color card.
  • the shape correction process may be to perform perspective transformation on the image, so that the image is mapped into a standard shape, wherein the standard shape is the shape of a gradient color card. For example, if each color card in the gradient color card is a rectangle, the standard shape is a rectangle; or, if each color card in the gradient color card is a parallelogram, the standard shape is a parallelogram.
  • the four vertex coordinates ⁇ x 1 ,x 2 ,x 3 ,x 4 , ⁇ of the gradient color card in the image can be corrected to obtain new four vertex coordinates ⁇ y 1 ,y 2 ,y 3 ,y 4 , ⁇ , and the shape enclosed by the new four vertex coordinates is a standard rectangle. Then according to the corresponding relationship between ⁇ x 1 ,x 2 ,x 3 ,x 4 , ⁇ and ⁇ y 1 ,y 2 ,y 3 ,y 4 , ⁇ , calculate the mapping relationship of the perspective transformation of the gradient color card, such as perspective transformation Matrix H k etc.
  • the shape correction processing of the present application may also be other mapping transformations with higher degrees of freedom, which are not limited in this application.
  • Image distortion caused by the shooting angle can be eliminated by shape correction processing.
  • the inconsistency of the magnification of the actual optical lens at different positions will also bring about imaging distortion, which makes the originally straight lines appear curved in the captured image, or the originally parallel lines are no longer parallel in the image.
  • the image may also be subjected to distortion correction to correct the imaging distortion of the image sensor.
  • the corrected image is then subjected to shape correction processing.
  • Those skilled in the art can select a correction algorithm to correct the imaging distortion according to actual needs, which is not discussed in this application.
  • the correction process is performed to obtain a corrected image that matches the shape of the gradient color card
  • at least one pixel point with different pixel values is extracted from the corrected image as a color sample set.
  • the set of color samples includes at least one color sample, which is a color output value of the image sensor.
  • the color sample may be the pixel value of one pixel, or may be the pixel value of multiple pixels.
  • the multiple pixel points may be multiple pixel points with the same pixel value, or may be multiple pixel points with similar pixel values. If the plurality of pixels are pixels with similar pixel values, the similar pixels may be smoothed first to obtain the same pixel value, and the color sample is the pixel value obtained after smoothing.
  • the color reference value may be a preset standard value, that is, a color standard value.
  • the preset standard value may be obtained based on a gradient model, and the gradient model is generated by training using a gradient color card. According to the gradient model of the gradient color card, the color standard value corresponding to any pixel point in the gradient color card can be calculated.
  • the preset standard value can also be obtained by interpolation calculation according to the pixel value at the preset position of the gradient color card.
  • the gradient color card T234 shown in Figure 3 includes 10 color bars, arranged in 5 rows and 2 columns.
  • the pixel values of the pixels in the same column of each color bar are the same.
  • the Lab space values corresponding to the pixels in the leftmost column and the rightmost column in the first color bar in the upper left corner are (76.77, -12.82, 37.92) and (46.59, -31.51, 24.03) respectively.
  • an interpolation algorithm such as linear interpolation, to fit a gradient model that satisfies the values of the two endpoints above, and then calculate the color standard value corresponding to any pixel point in the gradient color card from the gradient model.
  • the preset position of the gradient color card can be the pixel value of the pixel point at other positions in addition to the endpoint value of the gradient color card, and the pixel value can be expressed in the RGB space or the YUV space in addition to the Lab space. This does not limit. It is worth noting that under different light sources, the pixel values of the preset positions are different, so the gradient model under different light sources will also be different.
  • the color reference value may also be a color output value corresponding to a color sample by other image sensors.
  • the color output values corresponding to the color samples by other image sensors can be obtained through the steps shown in FIG. 4 :
  • Step 410 Obtain the second image of the gradient color card captured by the other image sensors
  • Step 420 Perform shape correction processing on the second image to obtain a second corrected image, so that the shape of the second corrected image matches the shape of the gradient color card;
  • Step 430 Extract a second color sample set consistent with the color sample set from the second corrected image, and acquire color output values corresponding to each color sample in the second color sample set.
  • steps 410 and 420 are similar to the specific implementation of the above-mentioned captured image and shape correction processing, and are not repeated in this application.
  • the image captured in step 210 and the second image in step 410 should be captured under the same light source condition.
  • a second color sample set consistent with the above-mentioned color sample set needs to be extracted therefrom, that is, the color output value corresponding to each color sample in the above-mentioned color sample set by other image sensors is used as the color reference value.
  • a color correction model can be generated based on the color samples and the color reference values, so as to perform color calibration on the image sensor.
  • the color sample and the color reference value corresponding to the color sample may be input into a specified neural network model for training, and the trained neural network model may be used as the color correction model.
  • the neural network model may be an RBF network, a convolutional neural network, or the like.
  • the color correction model can also be a color correction matrix (CCM, Color Correction Matrix), 3D LUT and other models.
  • CCM Color Correction Matrix
  • 3D LUT 3D LUT
  • the trained color correction model can be loaded in a camera or other electronic device, so that in the subsequent actual use of the camera or other electronic device, the image captured by the image sensor is input into the trained color correction model model, the color-corrected image can be output.
  • the gradient color card includes at least one gradient color bar corresponding to different color spaces.
  • the gradient color card T234 shown in Figure 3 includes 10 color bars, arranged in 5 rows and 2 columns.
  • the method further includes extracting at least one gradient color bar from the image captured by the image sensor, and extracting color samples from different gradient color bars.
  • each gradient color bar can be obtained by manual selection or automatic calculation to extract the gradient color bar.
  • shape correction processing needs to be performed on the image, including shape correction processing for each extracted gradient color bar, so that each corrected gradient color bar matches the shape of the gradient color card.
  • the specific shape correction process is as described above, which is not repeated in this application.
  • color samples can be extracted from different gradient color bars. Since different gradient color bars correspond to different color spaces, color samples are obtained from each gradient color bar respectively, and the formed color sample set has a wide distribution range. Evenly distributed and more representative without clustering in one part of the color space.
  • the present application provides a color calibration method. Since the colors in the gradient color are uniformly distributed in the entire color space, by extracting at least one pixel point with different pixel values from the gradient color card as a color sample, a large number of uniformly distributed pixels can be obtained. And representative color samples and their color reference values are used to train the color correction model, which greatly improves the model accuracy and color restoration accuracy. Among them, since the color reference value is obtained by interpolation calculation according to the pixel value of the preset position of the gradient color card, or obtained from the color output value of other image sensors, the position information of the pixel point has an important influence on the accuracy of the obtained color reference value.
  • the color reference value is the preset standard value, that is, the color standard value
  • the color calibration of the camera can be realized, so that the camera output color reproduction accuracy is high;
  • the color reference value is the color output value of other image sensors, different Alignment of image sensor color output values.
  • the color calibration method provided in this application can also perform color calibration under various light sources. Since the color temperatures corresponding to different light sources are different, color correction models corresponding to different light sources (color temperatures) can be established respectively. For example, the color temperature corresponding to the standard light source D65 is 6500K, and the color temperature corresponding to D50 is 5000K, then the images of the gradient color cards taken in sequence under D65 and D50 can be obtained respectively, and established according to the color calibration method described in any of the above embodiments. Corresponds to the color correction model of D65 and D50.
  • the interpolation algorithm or the weighted average can be used to adjust the contribution of the two models to the final output result. Color calibration results for images at any color temperature.
  • the above embodiment provides a color calibration method based on a gradient color card. In some embodiments, it is not limited to obtaining the training samples required for training the color correction model only through the gradient color card.
  • the color standard value of the standard color card is more reliable, it is necessary to properly adjust the color samples extracted from the gradient color card and the color samples extracted from the standard color card for color correction. Contribution of model training. For example, when the number of color samples extracted from the gradient color card is large, the weight of the color samples extracted from the standard color card can be increased to make the contribution of two color samples to the training of the color correction model. Equivalent, or can appropriately reduce the number of color samples extracted from the gradient color card. Those skilled in the art can adjust the contribution degrees of the color samples extracted from the gradient color card and the standard color card respectively according to actual needs, which is not limited in this application.
  • the standard color card can provide a limited number of training samples, and only represent a small number of colors in the color space.
  • the color standard value of the standard color card is more reliable, it can be combined with the standard color card and the gradient color card. By extracting a large number of uniformly distributed color samples from the gradient color card to cover the color space that cannot be covered by the standard color card, the accuracy, accuracy, and generalization ability of the color correction model can be effectively improved.
  • the present application also provides a color calibration method, as shown in Figure 5, including the steps:
  • Step 510 Acquire the image of the gradient color card captured by the image sensor
  • Step 520 Perform distortion correction on the image captured by the image sensor to correct the imaging distortion of the image sensor;
  • Step 530 Extract at least one gradient color bar from the distortion-corrected image, where the at least one gradient color bar corresponds to different color spaces;
  • Step 540 Perform shape correction processing on each extracted gradient color bar to obtain a corrected gradient color bar, wherein the shape of each corrected gradient color bar matches the shape of the gradient color card
  • Step 550 Smooth the at least one gradient color bar
  • Step 560 Extract at least one pixel point with different pixel values from the at least one gradient color bar as a color sample set, and obtain the color reference value corresponding to each color sample in the color sample set;
  • Step 570 Generate a color correction model based on the color sample and the color reference value to perform color calibration on the image sensor.
  • the present application also provides a schematic structural diagram of the controller as shown in FIG. 6 .
  • the controller includes a processor and a non-volatile memory; the processor is used to execute program instructions in the memory to implement the above solution, optionally, the controller may also include an internal bus, a network interface , one or more of the memory, and of course, may also include hardware required by other businesses.
  • the processor reads the corresponding computer program from the non-volatile memory into the memory and runs it, so as to implement the color calibration method described in any of the above embodiments.
  • the controller can be mounted in the camera, for example, the controller is an image processor (or controller or processor) in the camera, and the image processor can implement the present invention in addition to the general image processing process The color calibration scheme mentioned in the application; or the controller is another device independent from the image processor in the camera, which can be used to implement the color calibration scheme mentioned in this application; or the controller is another device connected to the camera A controller or processor on a device.
  • the controller is an image processor (or controller or processor) in the camera, and the image processor can implement the present invention in addition to the general image processing process The color calibration scheme mentioned in the application; or the controller is another device independent from the image processor in the camera, which can be used to implement the color calibration scheme mentioned in this application; or the controller is another device connected to the camera A controller or processor on a device.
  • the present application also provides a schematic structural diagram of a camera as shown in FIG. 7 .
  • the controller includes an image sensor and a controller.
  • the controller includes a processor and a non-volatile memory; the processor is used to execute program instructions in the memory to realize the above Solution, optionally, the controller may also include one or more of an internal bus, a network interface, and a memory, and of course, may also include hardware required by other services.
  • the processor reads the corresponding computer program from the non-volatile memory into the memory and executes it, so as to implement the color calibration method described in any of the foregoing embodiments.
  • the present application also provides a schematic structural diagram of an electronic device as shown in FIG. 8 .
  • the controller includes a controller.
  • the controller includes a processor and a non-volatile memory; the processor is used to execute program instructions in the memory to implement the above solution, and can
  • the controller may also include one or more of an internal bus, a network interface, and a memory, and of course, may also include hardware required by other services.
  • the processor reads the corresponding computer program from the non-volatile memory into the memory and executes it, so as to implement the color calibration method described in any of the foregoing embodiments.
  • the electronic device is connected to the camera, and is used to perform color calibration on the camera, for example, a parameter adjustment tool of the camera and the like.
  • the present application also provides a computer storage medium, where a computer program is stored in the storage medium, and when the computer program is executed by a processor, the computer program can be used to execute the color calibration method described in any of the foregoing embodiments.
  • controller embodiment since it basically corresponds to the method embodiment, it is sufficient to refer to the partial description of the method embodiment for related parts.
  • the controller embodiments described above are only illustrative, wherein the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, Located in one place, or it can be distributed over multiple network elements. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution in this embodiment. Those of ordinary skill in the art can understand and implement it without creative effort.

Landscapes

  • Engineering & Computer Science (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • Image Processing (AREA)

Abstract

本申请提供了一种颜色标定的方法、控制器、相机、电子设备及存储介质，所述方法包括：获取图像传感器拍摄的渐变色卡的图像；对所述图像进行形状校正处理，得到校正图像，以使所述校正图像的形状与所述渐变色卡的形状匹配；从所述校正图像提取至少一个不同像素值的像素点作为颜色样本集，并获取所述颜色样本集中各颜色样本对应的颜色基准值；由于从渐变色卡中可以提取代表色彩空间中大量颜色的颜色样本，因此基于颜色样本与颜色样本对应的基准值生成的颜色校正模型精度较高，从而提高了颜色还原精度。

Description

一种颜色标定方法、控制器、相机、电子设备及存储介质 技术领域
本申请涉及图像处理技术领域，尤其涉及一种颜色标定方法、控制器、相机、电子设备及存储介质。
背景技术
相机通过图像传感器，即感光元件记录所拍摄物体的色彩。由于不同图像传感器的光谱响应和对光电信号的处理算法不同，导致拍摄的物体在相机中呈现的颜色与人眼看到的颜色存在差异。为了还原人眼所感受到的颜色，相机在出厂前往往需要经过颜色标定，使得所拍摄的图像中物体呈现的颜色与其真实颜色一致。
传统颜色校正过程中校正精度较低，因此如何提高颜色校正的精准度是目前需要解决的技术问题。
发明内容
有鉴于此，本申请的目的之一是提供一种颜色标定方法、控制器、相机、电子设备及存储介质，可以提高颜色校正模型的精度即泛化能力，从而提高颜色还原精度。
为了达到上述技术效果，本申请实施例公开了如下技术方案：
第一方面，提供了一种颜色标定方法，所述方法包括：
获取图像传感器拍摄的渐变色卡的图像；
对所述图像进行形状校正处理，得到校正图像，以使所述校正图像的形状与所述渐变色卡的形状匹配；
从所述校正图像提取至少一个不同像素值的像素点作为颜色样本集，并获取所述颜色样本集中各颜色样本对应的颜色基准值；
基于所述颜色样本与所述颜色基准值生成颜色校正模型，以对所述图像传感器进行颜色标定。
第二方面，提供了一种控制器，所述控制器包括：
处理器；
用于存储处理器可执行指令的存储器；
所述处理器被配置为：
获取图像传感器拍摄的渐变色卡的图像；
对所述图像进行形状校正处理，得到校正图像，以使所述校正图像的形状与所述渐变色卡的形状匹配；
从所述校正图像提取至少一个不同像素值的像素点作为颜色样本集，并获取所述颜色样本集中各颜色样本对应的颜色基准值；
基于所述颜色样本与所述颜色基准值生成颜色校正模型，以对所述图像传感器进行颜色标定。
第三方面，提供了一种相机，所述相机包括图像传感器以及如第二方面所述的控制器。
第四方面，提供了一种电子设备，所述电子设备包括如第二方面所述的控制器，所述电子设备用于对相机进行颜色标定。
第五方面，提供了一种计算机存储介质，其上存储有计算机指令，该指令被处理器执行时实现上述任一的方法。
本申请的实施例提供的技术方案可以包括以下有益效果：
本申请提供了一种颜色标定方法、控制器、相机、电子设备及存储 介质，通过拍摄渐变色卡，并对拍摄图像进行形状校正处理后，从图像中提取至少一个不同像素值的像素点作为颜色样本集；由于从渐变色卡中可以提取代表色彩空间中大量颜色的颜色样本，因此基于颜色样本与颜色样本对应的基准值生成的颜色校正模型精度较高，从而提高了颜色还原精度。
附图说明
为了更清楚地说明本申请实施例中的技术方案，下面将对实施例描述中所需要使用的附图作简单地介绍，显而易见地，下面描述中的附图仅仅是本申请的一些实施例，对于本领域普通技术人员来讲，在不付出创造性劳动性的前提下，还可以根据这些附图获得其他的附图。
图1是相关技术中24色标准色卡和140色ColorChecker SG色卡。
图2是本申请根据一示例性实施例示出的一种颜色标定方法。
图3是本申请根据另一示例性实施例示出的TE234渐变色卡。
图4是本申请根据另一示例性实施例示出的一种颜色标定方法。
图5是本申请根据另一示例性实施例示出的一种颜色标定方法。
图6是本申请根据另一示例性实施例示出的一种控制器。
图7是本申请根据另一示例性实施例示出的一种相机。
图8是本申请根据另一示例性实施例示出的一种电子设备。
具体实施方式
下面将结合本申请实施例中的附图，对本申请实施例中的技术方案进行清楚、完整地描述，显然，所描述的实施例仅仅是本申请一部分实施例，而不是全部的实施例。
本申请中，“至少一个”是指一个或者多个，“多个”是指两个或两个以上。“和/或”，描述关联对象的关联关系，表示可以存在三种关系，例如，A和/或B，可以表示：单独存在A，同时存在A和B，单独存在B的情况，其中A、B可以是单数或者复数。字符“/”一般表示前后关联对象是一种“或”的关系。“以下至少一项(个)”或其类似表达，是指的这些项中的任意组合，包括单项(个)或复数项(个)的任意组合。例如，a、b或c中的至少一项(个)，可以表示：a、b、c，a和b，a和c，b和c，或a和b和c，其中a、b、c可以是单个，也可以是多个。
相机在出厂前或者用户在使用相机前往往需要先经过调参工具进行颜色标定，使得相机所拍摄的图像中物体呈现的颜色与真实颜色一致。通常地，相机在标准光源下拍摄24色标准色卡(图1左)或140色ColorChecker SG(图1右)等标准色卡。其中标准色卡包括多个色块，而每个色块在标准光源下的颜色标准值是已知的经验值，因此通过将各个色块的相机颜色输出值，以及色块对应的颜色标准值作为训练样本，可以对颜色校准模型进行训练，从而完成对相机颜色的标定。在之后的使用过程中，相机所拍摄的图像输入到训练好的颜色校准模型中进行颜色校准后，所输出的图像中物体呈现颜色可以与真实颜色一致。然而，常用的标准色卡颜色数量有限，如24色标准色卡只有24个色块，即能提供的训练样本数量只有24，而140色ColorChecker SG能提供的训练样本数量也只有140。常用的标准色卡能提供的训练样本数量不仅有限，且这些训练样本也仅代表颜色空间中的少量颜色。而当训练样本数量不足时，所训练的颜色校准模型精度不高，容易过拟合，导致模型的泛化能力较弱，具体表现为在相机的使用过程中，经过颜色校准模型校准后的图像颜色与真实颜色存在一定差别，且颜色过渡不自然。
为了解决上述技术问题，本申请提供了一种颜色标定方法，可以应用于相机、与相机相连的电子设备，以及其他搭载有图像传感器的电子设 备中。其中，相机包括图像传感器，即感光元件，用于记录所拍摄物体的色彩，而与相机相连的电子设备可以读取相机所存储的数据或参数，并能够修改相机参数或写入数据，如相机的调参工具。如图2所示，所述颜色标定方法包括：
步骤210：获取图像传感器拍摄的渐变色卡的图像；
步骤220：对所述图像进行形状校正处理，得到校正图像，以使所述校正图像的形状与所述渐变色卡的形状匹配；
步骤230：从所述校正图像提取至少一个不同像素值的像素点作为颜色样本集，并获取所述颜色样本集中各颜色样本对应的颜色基准值；
步骤240：基于所述颜色样本与所述颜色基准值生成颜色校正模型，以对所述图像传感器进行颜色标定。
本申请所提供的一种颜色标定方法，通过拍摄渐变色卡，并对拍摄图像进行形状校正处理后，从图像中提取至少一个不同像素值的像素点作为颜色样本集；由于从渐变色卡中可以提取代表色彩空间中大量颜色的颜色样本，因此基于颜色样本与颜色样本对应的基准值生成的颜色校正模型精度较高，从而提高了颜色还原精度。
其中，图像传感器可以在如D65、D50、TL84等标准光源下拍摄渐变色卡，渐变色卡在标准光源下的颜色标准值已知。除了使用标准光源，还可以使用其他任意颜色标准值已知的渐变色卡的光源进行拍摄，本申请在此不做限制。
在一些实施例中，渐变色卡可以是如图3所示的TE234渐变色卡，当然还可以是其他颜色标准值已知的商用或自定义渐变色卡，且不局限于彩色渐变色卡，也可以是灰度色卡，渐变色卡的形状也可以是多种，本申请在此不做限制。下面以如图3所示的渐变色卡TE234为例进行展述。
在拍摄过程中，拍摄的角度会导致所拍摄的图像发生形变。例如当 镜头与渐变色卡不平行时，在拍出来图像中，渐变色卡的形状可能会发生形变，例如渐变色卡为矩形，而拍摄出的图像中，变为平行四边，而非标准矩形。而图像的形状对后续步骤230中获取颜色样本对应的颜色基准值有重要的影响，因此需要先对图像进行形状校正处理，使得校正图像的形状与渐变色卡的形状匹配。
在一些实施例中，形状校正处理可以是对图像进行透视变换，使得图像映射为标准形状，其中，标准形状为渐变色卡的形状。例如，渐变色卡中每个色卡均为矩形，则标准形状为矩形；或者，渐变色卡中每个色卡均为平行四边形，则标准形状为平行四边形。以标准形状为矩形为例，可以对图像中渐变色卡的四个顶点坐标{x 1,x 2,x 3,x 4,}进行修正，得到新的四个顶点坐标{y 1,y 2,y 3,y 4,}，且新的四个顶点坐标所围成的形状为标准矩形。再根据{x 1,x 2,x 3,x 4,}与{y 1,y 2,y 3,y 4,}的对应关系，计算出渐变色卡的透视变换的映射关系，例如透视变换矩阵H k等。并通过上述映射关系对图像执行透视变换，使得图像中的渐变色卡的形状为标准矩形。当然若渐变色卡是其他形状的，则并不局限于通过透视变换将图像校正为标准矩形，而是校正为与渐变色卡形状匹配。此外，本申请的形状校正处理除了透视变换以外，还可以是其他自由度更高的映射变换，本申请在此不做限制。
通过形状校正处理，可以消除因拍摄角度导致的图像形变。但除了拍摄角度的影响之外，实际光学镜头不同位置的放大率不一致还会带来成像畸变，使得原本笔直的线条在拍摄图像中呈现弯曲，或原本平行的直线在图像中不再平行。为了修正成像畸变，在一些实施例中，在形状校正处理之前，还可以先对所述图像进行畸变校正，以对图像传感器的成像畸变进行修正。再将修正后的图像进行形状校正处理。本领域技术人员可以根据实际需要选取校正算法来修正成像畸变，本申请在此不展开论述。
在进行校正处理，得到与渐变色卡形状匹配的校正图像后，则从校 正图像提取至少一个不同像素值的像素点作为颜色样本集。颜色样本集包括至少一个颜色样本，颜色样本是图像传感器的颜色输出值。在一些实施例中，颜色样本可以是一个像素点的像素值，也可以是多个像素点的像素值。多个像素点可以是多个像素值相同的像素点，也可以是多个像素值相似的像素点。若多个像素点是多个像素值相似的像素点，则可以先将相似的像素点进行平滑处理，平滑成同一像素值，颜色样本为平滑后的得到的像素值。
进一步地，还需要获取颜色样本集中各颜色样本对应的颜色基准值。当然，在一些实施例中，获取所述颜色基准值之前，还可以先对校正图像进行平滑处理，包括但不限于执行平滑滤波，以减少图像噪声干扰。在一些实施例中，颜色基准值可以是预设的标准值，即颜色标准值。预设的标准值可以基于渐变模型获取，所述渐变模型是利用渐变色卡训练生成的。根据渐变色卡的渐变模型，可以计算渐变色卡内任意位置像素点对应的颜色标准值。预设的标准值还可以根据渐变色卡预设位置的像素值进行插值计算获取。举个例子，如图3所示的渐变色卡T234包括10个色条，按照5行2列排布。每个色条同一列的像素点像素值相同。在D50光源下，已知左上角第一个色条中最左列和最右列像素点对应的Lab空间取值分别为(76.77,-12.82,37.92)以及(46.59,-31.51,24.03)。则利用插值算法，如线性插值，拟合出满足上述两个端点取值的渐变模型，再从渐变模型中计算渐变色卡内任意位置像素点对应的颜色标准值。当然，上述渐变色卡预设位置除了渐变色卡的端点值还可以是其他位置像素点的像素值，且像素值除了以Lab空间表示外，还可以用RGB空间或YUV空间表示，本申请在此不做限制。值得注意的是，在不同光源下，预设位置的像素值不同，因此不同光源下的渐变模型也会不相同。
颜色基准值除了可以是预设的标准值，在一些实施例中，颜色基准值还可以是其他图像传感器对颜色样本对应的颜色输出值。通过将颜色样 本和其他图像传感器对颜色样本对应的颜色输出值，生成颜色校正模型，可以实现不同图像传感器之间输出颜色的对齐。其中，其他图像传感器可以是比当前图像传感器规格更高的传感器，例如图像分辨率、图像色彩、画质等比当前传感器更高的其他传感器。在一些实施例中，其他图像传感器对颜色样本对应的颜色输出值可以通过如图4所示的步骤获取：
步骤410：获取所述其他图像传感器拍摄的渐变色卡的第二图像；
步骤420：对所述第二图像进行形状校正处理，得到第二校正图像，以使所述第二校正图像的形状与所述渐变色卡的形状匹配；
步骤430：从所述第二校正图像提取与所述颜色样本集一致的第二颜色样本集，并获取所述第二颜色样本集中各颜色样本对应的颜色输出值。
其中，步骤410和步骤420与上述拍摄图像和形状校正处理的具体实施方式类似，本申请不再赘述。但步骤210中所拍摄的图像与步骤410中第二图像要在同一光源条件下拍摄。在得到第二校正图像后，需要从中提取与上述颜色样本集一致的第二颜色样本集，即以其他图像传感器对上述颜色样本集中各颜色样本对应的颜色输出值，作为颜色基准值。
在获取颜色样本及其对应的颜色基准值后，可以基于颜色样本与颜色基准值生成颜色校正模型，以对图像传感器进行颜色标定。在一些实施例中，可以将所述颜色样本和所述颜色样本对应的颜色基准值输入指定的神经网络模型进行训练，并将训练好的神经网络模型作为所述颜色校正模型。所述神经网络模型可以是RBF网络、卷积神经网络等。除了神经网络模型外，颜色校正模型还可以是颜色校正矩阵(CCM，Color Correction Matrix)、3D LUT等模型，本领域技术人员可以根据实际需要选取不同的模型作为训练模型，并将训练后的模型作为颜色校正模型，本申请在此不做限制。在一些实施例中，训练好的颜色校正模型可以被装载于相机或其他电子设备中，以使在后续的相机或其他电子设备实际使用过程中，图像 传感器所拍摄的图像输入训练好的颜色校正模型中，即可输出颜色校正后的图像。
在一些实施例中，渐变色卡包括至少一个对应于不同颜色空间的渐变色条。如图3所示的渐变色卡T234包括10个色条，按照5行2列排布。则在对图像进行形状校正处理，得到校正图像之前，还包括从所述图像传感器拍摄的图像中提取至少一个渐变色条，而颜色样本则从不同的渐变色条中提取。具体地，渐变色卡中第N个渐变色条X n可用其四个顶点坐标表示，记为X n＝{x n1,x n2,x n3,x n4,}。通过手动选取或自动计算的方式可以得到每个渐变色条的四个顶点，以提取渐变色条。在提取至少一个渐变色条后，需要对图像进行形状校正处理，包括对所提取的每个渐变色条进行形状校正处理，以使校正后的每个渐变色条与渐变色卡的形状匹配。具体的形状校正过程如上所述，本申请在此不再赘述。进一步地，颜色样本可以从不同的渐变色条中提取，由于不同的渐变色条对应着不同的颜色空间，因此从各个渐变色条中分别获取颜色样本，所组成的颜色样本集分布范围广且分布均匀，更具代表性，而不会聚集在色彩空间中的某一部分。
本申请提供了一种颜色标定方法，由于渐变色中颜色均匀分布在整个颜色空间中，因此通过从渐变色卡中提取至少一个不同像素值的像素点作为颜色样本，可以获取到大量且分布均匀并具有代表性的颜色样本及其颜色基准值来训练颜色校正模型，大大提高了模型精度以及颜色还原精度。其中，由于颜色基准值是根据渐变色卡预设位置的像素值进行插值计算获取，或从其他图像传感器的颜色输出值获取，像素点的位置信息对所获取的颜色基准值精度有重要的影响，为了对像素点进行精准定位，需要对图像进行形状校正处理，以确保所获取的颜色基准值的准确性。当颜色基准值为预设的标准值，即颜色标准值时，可以实现对相机的颜色校准，使得相机输出颜色还原精度高；当颜色基准值为其他图像传感器的颜色输出值时，可以实现不同图像传感器颜色输出值的对齐。
此外，本申请所提供的颜色标定方法，还可以在多种光源下进行颜色标定。由于不同光源对应的色温不一样，因此可以分别建立对应于不同光源(色温)的颜色校正模型。例如标准光源D65对应的色温为6500K，D50对应的色温为5000K，则可以分别获取在D65和D50下依次拍摄的渐变色卡的图像，并根据上述任一实施例所述的颜色标定方法分别建立对应D65和D50的颜色校正模型。在后续的使用过程中，若拍摄环境或光源对应的色温为5000K到6000K中任一色温，则可以通过插值算法，或通过加权平均，以调整两个模型对最终输出结果的贡献度，来得到任意色温下图像的颜色校准结果。
上述实施例提供了一种基于渐变色卡的颜色标定方法，在一些实施例中，并不局限于仅通过渐变色卡获取训练颜色校正模型所需的训练样本，还可以在渐变色卡的基础上结合常用的24色标准色卡和/或140色ColorChecker SG等标准色卡来获取颜色样本。由于在图像颜色检测领域中广泛使用24色标准色卡和140色ColorChecker SG等标准色卡，其颜色标准值的可信度更高，因此将渐变色卡与标准色卡结合使用，部分颜色样本从标准色卡中提取，可以使所训练出的颜色校正模型准确性和可信度更高。进一步地，在一些实施例中，由于标准色卡的颜色标准值可信度更高，因此需要适当地调整从渐变色卡中提取的颜色样本以及从标准色卡中提取的颜色样本对颜色校正模型训练的贡献度，例如，当从渐变色卡中提取的颜色样本数量较多时，可以通过增加从标准色卡中提取的颜色样本的权重，使得两种颜色样本对颜色校正模型训练的贡献度相当，又或者可以适当减少从渐变色卡提取颜色样本的数量。本领域技术人员可以根据实际需要调整分别从渐变色卡与标准色卡提取的颜色样本的贡献度，本申请在此不做限制。如上所述，标准色卡能提供的训练样本数量有限，且仅代表颜色空间中少量颜色，但由于标准色卡的颜色标准值可信度更高，因此可以结合标准色卡与渐变色卡，通过从渐变色卡中补充提取大量、分布均匀颜色样 本，以覆盖标准色卡未能覆盖的颜色空间，可以有效提高颜色校正模型的精确度、准确度、以及泛化能力。
此外，本申请还提供了一种颜色标定方法，如图5所示，包括步骤：
步骤510：获取图像传感器拍摄的渐变色卡的图像；
步骤520：对所述图像传感器拍摄的图像进行畸变校正，以对所述图像传感器的成像畸变进行修正；
步骤530：从畸变校正后的图像中提取至少一个渐变色条，所述至少一个渐变色条对应不同的颜色空间；
步骤540：对提取的每个渐变色条进行形状校正处理，得到校正后的渐变色条，其中，校正后的每个渐变色条的形状与所述渐变色卡的形状匹配
步骤550：对所述至少一个渐变色条进行平滑处理；
步骤560：从所述至少一个渐变色条中提取至少一个不同像素值的像素点作为颜色样本集，并获取所述颜色样本集中各颜色样本对应的颜色基准值；
步骤570：基于所述颜色样本与所述颜色基准值生成颜色校正模型，以对所述图像传感器进行颜色标定。
上述步骤的具体实现方式参见上文实施例，本申请在此不再赘述。
基于上述任意实施例所述的颜色标定方法，本申请还提供了如图6所示的控制器的结构示意图。如图6，在硬件层面，该控制器包括处理器、非易失性存储器；处理器用于执行存储器中的程序指令来实现上述方案，可选的，该控制器还可以包括内部总线、网络接口、内存中的一种或多种，当然还可能包括其他业务所需要的硬件。处理器从非易失性存储器中读取 对应的计算机程序到内存中然后运行，以实现上述任意实施例所述的颜色标定方法。其中，所述控制器可以搭载在相机中，例如，该控制器为相机中的图像处理器(或者控制器或者处理器)，该图像处理器除了执行通用的图像处理过程以外，还可以实现本申请提及的颜色标定方案；或者该控制器是与相机中的图像处理器独立的另外一个器件，可用于实现本申请提及的颜色标定方案；或者该控制器为与所述相机连接的其他设备上的控制器或者处理器。
基于上述任意实施例所述的颜色标定方法，本申请还提供了如图7所示的相机的结构示意图。如图7，在硬件层面，该控制器包括图像传感器和控制器，所述控制器如图6所示，包括处理器、非易失性存储器；处理器用于执行存储器中的程序指令来实现上述方案，可选的，该控制器还可以包括内部总线、网络接口、内存中的一种或多种，当然还可能包括其他业务所需要的硬件。处理器从非易失性存储器中读取对应的计算机程序到内存中然后运行，以实现上述任意实施例所述的颜色标定方法。
基于上述任意实施例所述的颜色标定方法，本申请还提供了如图8所示的电子的结构示意图。如图8，在硬件层面，该控制器包括控制器，所述控制器如图6所示，包括处理器、非易失性存储器；处理器用于执行存储器中的程序指令来实现上述方案，可选的，该控制器还可以包括内部总线、网络接口、内存中的一种或多种，当然还可能包括其他业务所需要的硬件。处理器从非易失性存储器中读取对应的计算机程序到内存中然后运行，以实现上述任意实施例所述的颜色标定方法。所述电子设备与相机相连，用于对所述相机进行颜色标定，例如可以是相机的调参工具等。
本申请还提供了一种计算机存储介质，存储介质存储有计算机程序，计算机程序被处理器执行时可用于执行上述任意实施例所述的颜色标定方法。
对于控制器实施例而言，由于其基本对应于方法实施例，所以相关 之处参见方法实施例的部分说明即可。以上所描述的控制器实施例仅仅是示意性的，其中所述作为分离部件说明的单元可以是或者也可以不是物理上分开的，作为单元显示的部件可以是或者也可以不是物理单元，即可以位于一个地方，或者也可以分布到多个网络单元上。可以根据实际的需要选择其中的部分或者全部模块来实现本实施例方案的目的。本领域普通技术人员在不付出创造性劳动的情况下，即可以理解并实施。
需要说明的是，在本文中，诸如第一和第二等之类的关系术语仅仅用来将一个实体或者操作与另一个实体或操作区分开来，而不一定要求或者暗示这些实体或操作之间存在任何这种实际的关系或者顺序。术语“包括”、“包含”或者其任何其他变体意在涵盖非排他性的包含，从而使得包括一系列要素的过程、方法、物品或者设备不仅包括那些要素，而且还包括没有明确列出的其他要素，或者是还包括为这种过程、方法、物品或者设备所固有的要素。在没有更多限制的情况下，由语句“包括一个……”限定的要素，并不排除在包括所述要素的过程、方法、物品或者设备中还存在另外的相同要素。
以上对本申请实施例所提供的方法和控制器进行了详细介绍，本文中应用了具体个例对本申请的原理及实施方式进行了阐述，以上实施例的说明只是用于帮助理解本申请的方法及其核心思想；同时，对于本领域的一般技术人员，依据本申请的思想，在具体实施方式及应用范围上均会有改变之处，综上所述，本说明书内容不应理解为对本申请的限制。

Claims (27)

  1. 一种颜色标定方法，其特征在于，所述方法包括：
    获取图像传感器拍摄的渐变色卡的图像；
    对所述图像进行形状校正处理，得到校正图像，以使所述校正图像的形状与所述渐变色卡的形状匹配；
    从所述校正图像提取至少一个不同像素值的像素点作为颜色样本集，并获取所述颜色样本集中各颜色样本对应的颜色基准值；
    基于所述颜色样本与所述颜色基准值生成颜色校正模型，以对所述图像传感器进行颜色标定。
  2. 根据权利要求1所述的方法，其特征在于，还包括步骤：将所述颜色样本和所述颜色样本对应的颜色基准值输入指定的神经网络模型进行训练，将训练好的神经网络模型作为所述颜色校正模型。
  3. 根据权利要求2所述的方法，其特征在于，所述指定的神经网络模型被装载于相机或其他电子设备中。
  4. 根据权利要求1所述的方法，其特征在于，所述颜色基准值为预设的标准值。
  5. 根据权利要求4所述的方法，其特征在于，所述预设的标准值基于渐变模型获取，所述渐变模型是利用所述渐变色卡训练生成的。
  6. 根据权利要求4所述的方法，其特征在于，所述预设的标准值根据所述渐变色卡预设位置的像素值进行插值计算获取。
  7. 根据权利要求1所述的方法，其特征在于，所述颜色基准值为其他图像传感器对所述颜色样本对应的颜色输出值。
  8. 根据权利要求7所述的方法，其特征在于，所述其他图像传感器对所述颜色样本对应的颜色输出值通过以下步骤获取：
    获取所述其他图像传感器拍摄的渐变色卡的第二图像；
    对所述第二图像进行形状校正处理，得到第二校正图像，以使所述第二校正图像的形状与所述渐变色卡的形状匹配；
    从所述第二校正图像提取与所述颜色样本集一致的第二颜色样本集，并获取所述第二颜色样本集中各颜色样本对应的颜色输出值。
  9. 根据权利要求1所述的方法，其特征在于，所述对所述图像进行形状校正处理，得到校正图像之前，所述方法还包括：从所述图像传感器拍摄的图像中提取至少一个渐变色条，所述至少一个渐变色条对应不同的颜色空间，所述颜色样本从所述渐变色条中提取；
    所述对所述图像进行形状校正处理，得到校正图像，包括：
    对提取的每个渐变色条进行形状校正处理，得到校正后的渐变色条，其中，校正后的每个渐变色条的形状与所述渐变色卡的形状匹配。
  10. 根据权利要求1所述的方法，其特征在于，所述对所述图像进行形状校正处理，得到校正图像之前，所述方法还包括：
    对所述图像传感器拍摄的图像进行畸变校正，以对所述图像传感器的成像畸变进行修正。
  11. 根据权利要求1所述的方法，其特征在于，在获取所述颜色基准值之前，还包括对所述校正图像进行平滑处理。
  12. 一种控制器，其特征在于，所述控制器包括：
    处理器；
    用于存储处理器可执行指令的存储器；
    所述处理器被配置为：
    获取图像传感器拍摄的渐变色卡的图像；
    对所述图像进行形状校正处理，得到校正图像，以使所述校正图像的形状与所述渐变色卡的形状匹配；
    从所述校正图像提取至少一个不同像素值的像素点作为颜色样本集，并获取所述颜色样本集中各颜色样本对应的颜色基准值；
    基于所述颜色样本与所述颜色基准值生成颜色校正模型，以对所述图像传感器进行颜色标定。
  13. 根据权利要求12所述的控制器，其特征在于，所述处理器用于： 将所述颜色样本和所述颜色样本对应的颜色基准值输入指定的神经网络模型进行训练，将训练好的神经网络模型作为所述颜色校正模型。
  14. 根据权利要求13所述的控制器，其特征在于，所述指定的神经网络模型被装载于相机或其他电子设备中。
  15. 根据权利要求12所述的控制器，其特征在于，所述颜色基准值为预设的标准值。
  16. 根据权利要求15所述的控制器，其特征在于，所述预设的标准值基于渐变模型获取，所述渐变模型是利用所述渐变色卡训练生成的。
  17. 根据权利要求15所述的控制器，其特征在于，所述预设的标准值根据所述渐变色卡预设位置的像素值进行插值计算获取。
  18. 根据权利要求12所述的控制器，其特征在于，所述颜色基准值为其他图像传感器对所述颜色样本对应的颜色输出值。
  19. 根据权利要求18所述的控制器，其特征在于，所述处理器用于：
    获取所述其他图像传感器拍摄的渐变色卡的第二图像；
    将所述第二图像进行校正处理，得到第二校正图像，以使所述第二校正图像的形状与所述渐变色卡的形状匹配；
    从所述第二校正图像中提取与所述颜色样本集一致的第二颜色样本集，并获取所述第二颜色样本集中各颜色样本对应的颜色输出值。
  20. 根据权利要求12所述的控制器，其特征在于，所述处理器用于：
    从所述图像传感器拍摄的图像中提取至少一个渐变色条，所述至少一个渐变色条对应不同的颜色空间，所述颜色样本从所述渐变色条中提取；
    对提取的每个渐变色条进行形状校正处理，得到校正后的渐变色条，其中，校正后的每个渐变色条的形状与所述渐变色卡的形状匹配。
  21. 根据权利要求12所述的控制器，其特征在于，所述处理器用于：
    对所述图像传感器拍摄的图像进行畸变校正，以对所述图像传感器的成像畸变进行修正。
  22. 根据权利要求12所述的控制器，其特征在于，在获取所述颜色基 准值之前，还包括对所述校正图像进行平滑处理。
  23. 根据权利要求12所述的控制器，其特征在于，所述控制器搭载在相机中，或与所述相机连接。
  24. 根据权利要求12所述的控制器，其特征在于，所述控制器为相机中的控制器，或者所述控制器为与相机连接的其他设备上的控制器。
  25. 一种相机，其特征在于，所述相机包括图像传感器以及如权利要求12所述的控制器。
  26. 一种电子设备，其特征在于，所述电子设备包括如权利要求12所述的控制器，所述电子设备用于对相机进行颜色标定。
  27. 一种计算机存储介质，其特征在于，其上存储有计算机指令，该指令被处理器执行时实现权利要求1至11任意一项所述的方法。
PCT/CN2021/076275 2021-02-09 2021-02-09 一种颜色标定方法、控制器、相机、电子设备及存储介质 WO2022170482A1 (zh)

Priority Applications (1)

Application Number Priority Date Filing Date Title
PCT/CN2021/076275 WO2022170482A1 (zh) 2021-02-09 2021-02-09 一种颜色标定方法、控制器、相机、电子设备及存储介质

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
PCT/CN2021/076275 WO2022170482A1 (zh) 2021-02-09 2021-02-09 一种颜色标定方法、控制器、相机、电子设备及存储介质

Publications (1)

Publication Number Publication Date
WO2022170482A1 true WO2022170482A1 (zh) 2022-08-18

Family

ID=82838070

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2021/076275 WO2022170482A1 (zh) 2021-02-09 2021-02-09 一种颜色标定方法、控制器、相机、电子设备及存储介质

Country Status (1)

Country Link
WO (1) WO2022170482A1 (zh)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN117249904A (zh) * 2023-11-15 2023-12-19 深圳市宗匠科技有限公司 颜色传感器的标定方法、装置、美容面罩及存储介质

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20080166043A1 (en) * 2007-01-05 2008-07-10 Silicon Optix Inc. Color and geometry distortion correction system and method
CN106023238A (zh) * 2016-06-30 2016-10-12 北京大学 一种相机模组颜色数据标定方法
CN108600723A (zh) * 2018-07-20 2018-09-28 长沙全度影像科技有限公司 一种全景相机的颜色校正方法及评价方法
CN112243584A (zh) * 2019-09-18 2021-01-19 深圳市大疆创新科技有限公司 图像的颜色校正方法、拍摄设备、图像的颜色校正系统

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20080166043A1 (en) * 2007-01-05 2008-07-10 Silicon Optix Inc. Color and geometry distortion correction system and method
CN106023238A (zh) * 2016-06-30 2016-10-12 北京大学 一种相机模组颜色数据标定方法
CN108600723A (zh) * 2018-07-20 2018-09-28 长沙全度影像科技有限公司 一种全景相机的颜色校正方法及评价方法
CN112243584A (zh) * 2019-09-18 2021-01-19 深圳市大疆创新科技有限公司 图像的颜色校正方法、拍摄设备、图像的颜色校正系统

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN117249904A (zh) * 2023-11-15 2023-12-19 深圳市宗匠科技有限公司 颜色传感器的标定方法、装置、美容面罩及存储介质
CN117249904B (zh) * 2023-11-15 2024-02-13 深圳市宗匠科技有限公司 颜色传感器的标定方法、装置、美容面罩及存储介质

Similar Documents

Publication Publication Date Title
CN111292246B (zh) 图像颜色校正方法、存储介质及内窥镜
US7346221B2 (en) Method and system for producing formatted data related to defects of at least an appliance of a set, in particular, related to blurring
Wang et al. Example-based image color and tone style enhancement
CN110021047A (zh) 图像处理方法、图像处理装置和存储介质
JP6455764B2 (ja) 色補正パラメータ算出方法、色補正パラメータ算出装置および画像出力システム
CN108600723A (zh) 一种全景相机的颜色校正方法及评价方法
EP3706042B1 (en) Image processing method, image processing apparatus, program, image processing system, and manufacturing method of learnt model
CN109003312B (zh) 一种基于非线性优化的相机标定方法
JP2000311243A (ja) 画像色補正方法及び装置
CN103268596B (zh) 一种降低图像噪声和使颜色接近标准的方法
CN108230407B (zh) 一种图像的处理方法和装置
CN110660034B (zh) 图像校正方法、装置及电子设备
CN108540716A (zh) 图像处理方法、装置、电子设备及计算机可读存储介质
JPH0646448A (ja) 画像処理方法および装置
CN110213556B (zh) 单纯色场景下的自动白平衡方法及系统、存储介质及终端
Vazquez-Corral et al. Color stabilization along time and across shots of the same scene, for one or several cameras of unknown specifications
WO2020093653A1 (zh) 色彩调整方法、色彩调整装置、电子设备及计算机可读存储介质
WO2022170482A1 (zh) 一种颜色标定方法、控制器、相机、电子设备及存储介质
CN110553739B (zh) 一种红外热成像的无挡片非均匀性校正方法
Grundmann et al. Post-processing approach for radiometric self-calibration of video
US20230098058A1 (en) System and method of white balancing a digital image with multiple light sources
CN111010557A (zh) 白点范围确定方法、白平衡校正方法及装置、存储介质
CN112601079B (zh) 一种摄像模组的校准方法、装置、设备及介质
CN115170670A (zh) 外参标定方法、装置及程序产品
KR101005625B1 (ko) 카메라 색상 특성 곡선을 이용한 영상 밝기 변화에 따른 색 보정 방법

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 21925162

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 21925162

Country of ref document: EP

Kind code of ref document: A1