WO2019047409A1 - 图像处理方法、系统、可读存储介质及移动摄像设备 - Google Patents
图像处理方法、系统、可读存储介质及移动摄像设备 Download PDFInfo
- Publication number
- WO2019047409A1 WO2019047409A1 PCT/CN2017/116362 CN2017116362W WO2019047409A1 WO 2019047409 A1 WO2019047409 A1 WO 2019047409A1 CN 2017116362 W CN2017116362 W CN 2017116362W WO 2019047409 A1 WO2019047409 A1 WO 2019047409A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- image
- pixel point
- current pixel
- pixel value
- pixel
- Prior art date
Links
- 238000003672 processing method Methods 0.000 title claims abstract description 27
- 230000002146 bilateral effect Effects 0.000 claims abstract description 41
- 238000001914 filtration Methods 0.000 claims abstract description 37
- 238000009877 rendering Methods 0.000 claims abstract description 26
- 238000000034 method Methods 0.000 claims abstract description 21
- 230000003796 beauty Effects 0.000 claims description 79
- 230000002087 whitening effect Effects 0.000 claims description 44
- 238000004590 computer program Methods 0.000 claims description 6
- 238000010606 normalization Methods 0.000 claims description 4
- 238000003384 imaging method Methods 0.000 description 8
- 238000006243 chemical reaction Methods 0.000 description 5
- 210000004709 eyebrow Anatomy 0.000 description 4
- 230000000694 effects Effects 0.000 description 3
- 210000004209 hair Anatomy 0.000 description 3
- 238000003491 array Methods 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 230000006870 function Effects 0.000 description 2
- 239000000463 material Substances 0.000 description 2
- 238000010586 diagram Methods 0.000 description 1
- 239000000835 fiber Substances 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/90—Determination of colour characteristics
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T5/00—Image enhancement or restoration
- G06T5/70—Denoising; Smoothing
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20024—Filtering details
- G06T2207/20028—Bilateral filtering
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20172—Image enhancement details
- G06T2207/20192—Edge enhancement; Edge preservation
Definitions
- the present invention relates to the field of image processing technologies, and in particular, to an image processing method, system, readable storage medium, and mobile imaging device.
- the current processing method for realizing the image beauty is to use the CPU to perform format conversion (usually conversion to RGB image) on the entire picture taken by the camera, and then use the CPU to perform filtering, skin color adjustment and the like,
- format conversion usually conversion to RGB image
- the filtering algorithm does not protect the surrounding environment and the boundary details of the human eyebrows, hair, etc., and the distortion is large.
- an object of the present invention is to provide an image processing method, system, readable storage medium, and mobile imaging apparatus with low distortion.
- the original RGB image is subjected to bilateral filtering and boundary filtering processing to obtain a bilateral filtered image and a boundary filtered image correspondingly;
- the pixel value of the current pixel in the original RGB image is used as the beauty pixel value corresponding to the current pixel point;
- the current pixel is on the original RGB image and the bilateral filtered image. Performing a mixing operation on the pixel values, and using the calculated mixed pixel value as the beauty pixel value corresponding to the current pixel point;
- an image processing method may further have the following additional technical features:
- the method further includes:
- W P is used for normalization
- I filtered (x) refers to the filtered image
- I(x) refers to the original image
- x i refers to the coordinates of the pixel point of the current filtering process
- ⁇ refers to the center point of the pixel point coordinate
- g s is The convolution kernel of space, where f r refers to the range kernel;
- the formula for mixing the pixel values of the current pixel point on the original RGB image and the bilaterally filtered image is: (1-smooth)*(A(x i , y i )-B(x i , y i ))+B(x i ,y i ),
- smooth is smoothness
- A(x i , y i ) is a pixel value of the current pixel point in the original RGB image
- B(x i , y i ) is the current pixel point on the bilateral Filter the pixel values on the image.
- the method further includes:
- the formula for performing the whitening operation on the beauty pixel value corresponding to the current pixel point is: w(x i , y i ) is the beauty pixel value corresponding to the current pixel point, and v(x i , y i ) is the corresponding white pixel value obtained after performing whitening operation on the current pixel point , ⁇ is the brightness enhancement coefficient.
- the step of determining whether the current pixel point is a pixel point of the skin color comprises:
- An image processing module is configured to perform bilateral filtering and boundary filtering processing on the original RGB image to obtain a bilateral filtered image and a boundary filtered image;
- a pixel value judging module configured to determine, when it is determined that any current pixel point in the original RGB image is a pixel point of a skin color, whether a pixel value of the current pixel point in the boundary filtered image is less than a threshold value;
- a first processing module configured to determine, when the pixel value of the current pixel in the boundary filtered image is not less than the threshold, a pixel value of the current pixel in the original RGB image as the The pixel value corresponding to the current pixel point;
- a second processing module configured to determine, when the pixel value of the current pixel in the boundary filtered image is less than the threshold, the current pixel point on the original RGB image and the bilateral filtered image Performing a mixing operation on the pixel values, and using the calculated mixed pixel value as the beauty pixel value corresponding to the current pixel point;
- a beauty rendering module configured to perform the beauty rendering of the original RGB image according to the beauty pixel value corresponding to each of the current pixel points.
- an image processing system may further have the following additional technical features:
- the image processing system further includes:
- a pixel point determining module configured to determine whether the current pixel point is a pixel point of the skin color
- the first processing module uses the pixel value of the current pixel point in the original RGB image as the beauty corresponding to the current pixel point. Pixel values.
- the image processing system further includes:
- a whitening operation module configured to perform a whitening operation on the beauty pixel value corresponding to the current pixel point to obtain a whitening pixel value corresponding to the current pixel point;
- a whitening rendering module configured to perform whitening rendering on the original RGB image according to the whitening pixel value corresponding to each of the current pixel points;
- the formula for performing the whitening operation on the beauty pixel value corresponding to the current pixel point is: w(x i , y i ) is the beauty pixel value corresponding to the current pixel point, and v(x i , y i ) is the corresponding white pixel value obtained after performing whitening operation on the current pixel point , ⁇ is the brightness enhancement coefficient.
- the present invention also provides a computer readable storage medium having stored thereon a computer program that, when executed by a processor, implements the image processing method described above.
- the present invention also provides a mobile imaging apparatus comprising a memory, a processor, and a computer program stored on the memory and operable on the processor, the processor being a GPU, the processor implementing the program to implement the above Image processing method.
- the image processing method, the system, the readable storage medium, and the mobile imaging device perform boundary filtering processing on the original RGB image to obtain the boundary filtered image, and then perform each pixel point in the original RGB image. Judging, when it is determined that the current pixel is a skin color and the pixel value of the boundary filtered image is not less than the threshold, and the pixel is a boundary pixel in the skin color, the beauty is rendered on the pixel Taking the original pixel value of the pixel in the original RGB image, when it is determined that the current pixel is the skin color and the pixel value of the boundary filtered image is smaller than the threshold, the pixel is represented as the skin color and For a non-boundary pixel, when the pixel is rendered in the face, the pixel value of the pixel in the original RGB image and the bilateral filtered image is used.
- the image processing method, system, readable storage medium, and mobile imaging device can better protect the boundary details of the eyebrows, hair, and the like while the skin color is beautiful, and the skin color is non-boundary
- the pixel points have been comprehensively considered, so the image distortion after the beauty is small, and it is more closely related to the real picture.
- FIG. 1 is a flow chart of an image processing method in a first embodiment of the present invention.
- FIG. 2 is a flow chart of an image processing method in a second embodiment of the present invention.
- FIG. 3 is a schematic structural diagram of an image processing system in a third embodiment of the present invention.
- Image processing module 11 Pixel point judgment module 12 First processing module 13 Second processing module 14 Beauty rendering module 15 Skin color judgment unit 121 Threshold judgment unit 122 Whitening module 16 Whitening rendering module 17 Image conversion module 18 Skin color judgment subunit 1211
- FIG. 1 there is shown an image processing method according to a first embodiment of the present invention, which includes steps S01 to S03.
- step S01 the original RGB image is subjected to bilateral filtering and boundary filtering processing to obtain a bilateral filtered image and a boundary filtered image.
- the purpose of processing the original RGB image by bilateral filtering is to preserve edge denoising, and the edge of the original RGB image is protected while processing the original RGB image.
- the boundary filtering process is also called Canny filtering processing, and the Canny filtering processing is implemented based on the Canny algorithm.
- the purpose of the Canny filtering processing is to find the boundary of the original RGB image.
- Step S02 When it is determined that any current pixel point in the original RGB image is a pixel point of the skin color, it is determined whether the pixel value of the current pixel point in the boundary filtered image is less than a threshold.
- the step S02 may sequentially determine each pixel point according to the arrangement order of the pixel points in the original RGB image.
- the pixel value of the skin color is within a certain range, by determining whether the pixel value of the pixel in the original RGB image is within the skin color (skin color pixel value), it can be determined whether the pixel is color.
- a pixel value in the original RGB image is not smaller than the threshold value in the boundary filtered image, it represents a portion where the pixel point is relatively sharp in the image, The pixel is a boundary.
- the pixel is uniformly changed in the image, and the pixel is non-boundary. .
- Step S03 is performed, when the step S02 determines that the current pixel point is in the boundary filter map.
- the pixel value in the image is smaller than the threshold, it represents that the pixel is a skin color and is a non-boundary pixel, and step S04 is performed.
- Step S03 the pixel value of the current pixel point in the original RGB image is taken as the beauty pixel value corresponding to the current pixel point.
- step S03 when the current pixel point is determined to be a boundary pixel point in the skin color, the original pixel value of the pixel point is directly used as a subsequent beauty rendering. Protects the boundaries of the skin tone.
- the step S03 is performed to better protect the surrounding environment (such as the background) in the original RGB image. ).
- Step S04 performing a mixing operation on the pixel values of the current pixel point on the original RGB image and the bilateral filtered image, and using the calculated mixed pixel value as the beauty pixel corresponding to the current pixel point. value.
- the process of performing the mixing operation on the pixel values of the current pixel point on the original RGB image and the bilaterally filtered image is: the pixel value of the current pixel point in the original RGB image and in the The pixel values on the bilateral filtered image are assigned a certain weight ratio, and then the mixing operation is performed to finally obtain a comprehensive pixel value, which is the beauty pixel value.
- the purpose of the step S04 is to realize that when the current pixel point is a skin color and is a non-boundary pixel point, the original pixel value of the pixel point and the bilaterally filtered pixel value are integrated. While beauty is applied to skin color and non-boundary pixels, try to make it fit to the real state, and the distortion is small.
- Step S05 Perform the beauty rendering of the original RGB image according to the beauty pixel value corresponding to each of the current pixel points.
- the beauty pixel value corresponding to each pixel in the original RGB image can be obtained, and then each of the beauty pixel values is rendered in a corresponding At the pixel point, the beauty of the original RGB image can be completed.
- the image processing method in the above embodiment of the present invention performs boundary filtering processing on the original RGB image to obtain the boundary filtered image, and then in the original RGB image. Each pixel point is judged. When it is determined that the current pixel point is not a skin color, the pixel point is a pixel point of the surrounding environment, and when the pixel point is rendered by the face, the pixel point is used in the original RGB.
- An original pixel value in the image when it is determined that the current pixel point is a skin color and the pixel value of the boundary filtered image is not less than the threshold value, and the pixel point is a boundary pixel point in the skin color, then the pixel is
- the point is used for the rendering of the face
- the original pixel value of the pixel in the original RGB image is adopted, and when it is determined that the current pixel is the skin color and the pixel value of the boundary filtered image is less than the threshold
- the pixel is a skin color and is a non-boundary pixel
- a mixed pixel value of the pixel value of the pixel in the original RGB image and the bilateral filtered image is used.
- the image processing method, system, readable storage medium, and mobile imaging device can better protect the surrounding environment (such as background), eyebrows, and hair while being beautiful for the skin color. Border detail features and complexion of the non-boundary pixels and a comprehensive consideration of beauty, it is a small image distortion after the beauty, the real picture is more appropriate.
- FIG. 2 an image processing method according to a second embodiment of the present invention is shown, which includes steps S11 to S18.
- step S11 the YUV image acquired by the camera is converted into a raw RGB image.
- the image acquired by the camera is usually a YUV image, and the acquired YUV image needs to be converted into a corresponding original RGB image to facilitate subsequent beauty operations.
- the existing YUV and RGB format conversion formula can be used to convert the YUV image acquired by the camera into the original RGB image.
- Step S12 Perform bilateral filtering and boundary filtering processing on the original RGB image to obtain a bilateral filtered image and a boundary filtered image.
- W P is used for normalization
- I filtered (x) refers to the filtered image
- I(x) refers to the original image
- x i refers to the coordinates of the pixel point of the current filtering process
- ⁇ refers to the center point of the pixel point coordinate
- g s is The convolution kernel of space
- f r refers to the range kernel.
- step S13 it is determined whether any current pixel point in the original RGB image is a pixel point of the skin color.
- the step of determining whether the current pixel point is the skin color pixel point includes:
- Determining whether a pixel value of the current pixel point in the original RGB image is within a range of pixel values of the skin color When it is determined that the pixel value of the current pixel in the original RGB image is within a pixel value range of the skin color, determining that the current pixel point is a pixel point of the skin color, that is, a skin color pixel point When it is determined that the pixel value of the current pixel in the original RGB image is not within the pixel value range of the skin color, determining that the current pixel point is not the pixel point of the skin color, that is, Non-skinned pixels.
- step S13 determines that the current pixel point is not the pixel point of the skin color, and represents the pixel point as the surrounding environment
- step S14, step S17, step S18, and step S19 are sequentially performed, when step S14 is performed.
- step S15 is performed.
- Step S14 the pixel value of the current pixel point in the original RGB image is used as the beauty pixel value corresponding to the current pixel point.
- Step S15 Determine whether a pixel value of the current pixel point in the boundary filtered image is less than a threshold.
- step S15 determines that the pixel value of the current pixel in the boundary filtered image is smaller than the threshold, and represents that the pixel is a skin color and is a non-boundary pixel
- step S16 to step S19 are performed.
- the step S15 determines that the pixel value of the current pixel in the boundary filtered image is not less than the threshold, then the pixel is represented as a boundary pixel in the skin color, and then the step S14 is performed.
- Step S16 performing a mixing operation on the pixel values of the current pixel point on the original RGB image and the bilateral filtered image, and using the calculated mixed pixel value as the beauty pixel corresponding to the current pixel point. value.
- the formula for mixing the pixel values of the current pixel point on the original RGB image and the bilaterally filtered image is: (1-smooth)*(A(x i , y i )-B(x i , y i )) + B(x i , y i ),
- A(x i , y i ) is the pixel value of the current pixel point in the original RGB image
- B(x i , y i ) is the pixel of the current pixel point on the bilateral filtered image Value
- smooth is smoothness, which can be adjusted according to actual conditions. When the smoothness is higher, the proportion of the bilaterally filtered image is larger.
- the smoothness is 1, the beauty pixel value of the current pixel point is directly Taking the pixel value of the current pixel on the bilateral filtered image, when the smoothness is 0, the beauty pixel value of the current pixel directly adopts the original pixel value on the original RGB image. .
- Step S17 Perform the beauty rendering of the original RGB image according to the beauty pixel value corresponding to each of the current pixel points.
- Step S18 Perform a whitening operation on the beauty pixel value corresponding to the current pixel point to obtain a whitening pixel value corresponding to the current pixel point.
- the formula for performing the whitening operation on the beauty pixel value corresponding to the current pixel point is: w(x i , y i ) is the beauty pixel value corresponding to the current pixel point, and v(x i , y i ) is the corresponding white pixel value obtained after performing whitening operation on the current pixel point , ⁇ is the brightness enhancement coefficient, which can be adjusted according to the actual situation.
- Step S19 Perform whitening rendering on the original RGB image according to the whitening pixel value corresponding to each of the current pixel points.
- the gradation of the original image is mapped to a slightly weaker end by using a log curve, and the middle is slightly stronger, and the brightness of the screen can be naturally enhanced to achieve a whitening effect.
- the image processing method in the above embodiment of the present invention compared with the image processing method in the first embodiment, after performing the beauty on the original RGB image, will also be the entire image after the beauty Whitening treatment is carried out to further enhance the image effect after the beauty.
- FIG. 3 is an image processing system according to a third embodiment of the present invention.
- the image processing system is a GPU, and the image processing system is disposed on a mobile camera.
- the device such as mobile phones, tablets, etc., including:
- the image processing module 11 is configured to perform bilateral filtering and boundary filtering processing on the original RGB image to Correspondingly, a bilateral filtered image and a boundary filtered image are obtained;
- the pixel value judging module 12 is configured to determine, when it is determined that any current pixel point in the original RGB image is a pixel point of the skin color, whether the pixel value of the current pixel point in the boundary filtered image is less than a threshold value;
- the first processing module 13 is configured to determine, when the pixel value in the boundary filtered image is not less than the threshold, the pixel value of the current pixel in the original RGB image Depicting the beauty pixel value corresponding to the current pixel point;
- the second processing module 14 is configured to determine, when the pixel value of the current pixel in the boundary filtered image is less than the threshold, the current pixel point on the original RGB image and the bilateral filtered image a pixel value is mixed, and the calculated mixed pixel value is taken as the beauty pixel value corresponding to the current pixel point;
- the beauty rendering module 15 is configured to perform the beauty rendering of the original RGB image according to the beauty pixel value corresponding to each of the current pixel points.
- W P is used for normalization
- I filtered (x) refers to the filtered image
- I(x) refers to the original image
- x i refers to the coordinates of the pixel point of the current filtering process
- ⁇ refers to the center point of the pixel point coordinate
- g s is The convolution kernel of space, where f r refers to the range kernel;
- the formula for mixing the pixel values of the current pixel point on the original RGB image and the bilaterally filtered image is: (1-smooth)*(A(x i , y i )-B(x i , y i ))+B(x i ,y i ),
- smooth is smoothness
- A(x i , y i ) is a pixel value of the current pixel point in the original RGB image
- B(x i , y i ) is the current pixel point on the bilateral Filter the pixel values on the image.
- the image processing system further includes:
- the pixel point judging module 16 is configured to determine whether the current pixel point is a pixel point of the skin color
- the first processing module 13 When it is determined that the current pixel point is not the pixel point of the skin color, the first processing module 13 The pixel value of the current pixel point in the original RGB image is used as the beauty pixel value corresponding to the current pixel point.
- the image processing system further includes:
- the whitening operation module 17 is configured to perform a whitening operation on the beauty pixel value corresponding to the current pixel point to obtain a white pixel value corresponding to the current pixel point;
- a whitening rendering module 18 configured to perform whitening rendering on the original RGB image according to the whitening pixel value corresponding to each of the current pixel points;
- the formula for performing the whitening operation on the beauty pixel value corresponding to the current pixel point is: w(x i , y i ) is the beauty pixel value corresponding to the current pixel point, and v(x i , y i ) is the corresponding white pixel value obtained after performing whitening operation on the current pixel point , ⁇ is the brightness enhancement coefficient.
- the image processing system further includes:
- the image conversion module 19 is configured to convert the YUV image acquired by the camera into the original RGB image.
- the pixel point judging module 16 includes:
- the pixel point judging sub-unit 161 is configured to determine whether a pixel value of each of the pixel points in the original RGB image is within a skin color range.
- the computing resources of the CPU can be better released and accelerated as compared with the prior art.
- the entire image processing process so that real-time beauty effects can also be achieved on the mobile platform.
- the present invention also provides a computer readable storage medium having stored thereon a computer program that, when executed by a processor, implements an image processing method as described above.
- the present invention also provides a mobile imaging device including a memory, a processor, and a computer program stored on the memory and operable on the processor, the processor being a GPU, and the processor implementing the program as described above Image processing method.
- a "computer-readable medium” can be any apparatus that can contain, store, communicate, propagate, or transport a program for use in an instruction execution system, apparatus, or apparatus, or in conjunction with such an instruction execution system, apparatus, or apparatus.
- computer readable media include the following: electrical connections (electronic devices) having one or more wires, portable computer disk cartridges (magnetic devices), random access memory (RAM), Read only memory (ROM), erasable editable read only memory (EPROM or flash memory), fiber optic devices, and portable compact disk read only memory (CDROM).
- the computer readable medium may even be a paper or other suitable medium on which the program can be printed, as it may be optically scanned, for example by paper or other medium, followed by editing, interpretation or, if appropriate, other suitable The method is processed to obtain the program electronically and then stored in computer memory.
- portions of the invention may be implemented in hardware, software, firmware or a combination thereof.
- multiple steps or methods may be implemented in software or firmware stored in a memory and executed by a suitable instruction execution system.
- a suitable instruction execution system For example, if implemented in hardware, as in another embodiment, it can be implemented by any one or combination of the following techniques well known in the art: having logic gates for implementing logic functions on data signals. Discrete logic circuits, application specific integrated circuits with suitable combinational logic gates, programmable gate arrays (PGAs), field programmable gate arrays (FPGAs), etc.
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Image Processing (AREA)
- Image Analysis (AREA)
Abstract
本发明提供一种图像处理方法、系统、可读存储介质及移动摄像设备，该方法包括：将原始RGB图像进行双边滤波及边界滤波处理，以得到双边滤波图像及边界滤波图像；当判断到原始RGB图像中的任意当前像素点为肤色的像素点时，判断当前像素点在边界滤波图像中的像素值是否小于阈值；若否，则将当前像素点在原始RGB图像中的像素值作为当前像素点对应的美颜像素值；若是，则将当前像素点在原始RGB图像及双边滤波图像上的像素值进行混合运算，并将运算出的混合像素值作为当前像素点对应的美颜像素值；根据每个当前像素点对应的美颜像素值，将原始RGB图像进行美颜渲染。本发明当中的图像处理方法，较好的保护了图像的边界，失真度小，美颜后更贴切于真实图片。
Description
本发明涉及图像处理技术领域，特别涉及一种图像处理方法、系统、可读存储介质及移动摄像设备。
随着摄像及图像处理技术的不断进步，人们在追求高像素的同时还要求所拍摄的图像能够自动美颜，即摄像设备能够自动对所拍图像进行美颜。为了满足广大用户的需要，许多电子产商都在所生产的移动摄像设备附加了自动美颜的功能。
然而，现有技术当中，目前实现图像美颜的处理方法为，利用CPU对摄像头拍摄的整张图片进行格式转换(通常转换为RGB图像)，然后再利用CPU进行滤波、肤色调整等运算，以达到自动美颜效果，整个美颜过程为串行运算，并且滤波算法没有保护周围环境以及人的眉毛、头发等边界细节特征，失真度大。
发明内容
基于此，本发明的目的是提供一种失真度小的图像处理方法、系统、可读存储介质及移动摄像设备。
根据本发明实施例的一种图像处理方法，包括：
将原始RGB图像进行双边滤波及边界滤波处理，以对应得到双边滤波图像及边界滤波图像；
当判断到所述原始RGB图像中的任意当前像素点为肤色的像素点时，判断所述当前像素点在所述边界滤波图像中的像素值是否小于阈值；
若否，则将所述当前像素点在所述原始RGB图像中的像素值作为所述当前像素点对应的美颜像素值；
若是，则将所述当前像素点在所述原始RGB图像及所述双边滤波图像上的
像素值进行混合运算，并将运算出的混合像素值作为所述当前像素点对应的所述美颜像素值；
根据每个所述当前像素点对应的所述美颜像素值，将所述原始RGB图像进行美颜渲染。
另外，根据本发明上述实施例的一种图像处理方法，还可以具有如下附加的技术特征：
进一步地，在所述将原始RGB图像进行双边滤波及边界滤波处理，以对应得到双边滤波图像及边界滤波图像的步骤之后，还包括：
判断所述当前像素点是否为所述肤色的像素点；
若否，则执行所述将所述当前像素点在所述原始RGB图像中的像素值作为所述当前像素点对应的美颜像素值的步骤。
进一步地，将所述原始RGB图像进行双边滤波处理的公式为：
其中，
WP用于归一化，Ifiltered(x)指滤波后的图像，I(x)指原始图像，xi指当前滤波处理的像素点的坐标，Ω指像素点坐标中心点，gs是空间的卷积核，fr指值域核；
将所述当前像素点在所述原始RGB图像及所述双边滤波图像上的像素值进行混合运算的公式为：(1-smooth)*(A(xi,yi)-B(xi,yi))+B(xi,yi)，
其中，smooth为光滑度，A(xi,yi)为所述当前像素点在所述原始RGB图像中的像素值，B(xi,yi)为所述当前像素点在所述双边滤波图像上的像素值。
进一步地，在所述根据每个所述当前像素点对应的所述美颜像素值，将所述原始RGB图像进行美颜渲染的步骤之后，还包括：
将所述当前像素点对应的所述美颜像素值进行美白运算，以得到所述当前像素点对应的美白像素值；
根据每个所述当前像素点对应的所述美白像素值，将所述原始RGB图像进
行美白渲染；
其中，所述将所述当前像素点对应的所述美颜像素值进行美白运算的公式为：
w(xi,yi)为所述当前像素点对应的所述美颜像素值，v(xi,yi)为所述当前像素点进行美白运算后得到的对应的所述美白像素值，β为亮度增强系数。
进一步地，所述判断所述当前像素点是否为所述肤色的像素点的步骤包括：
判断所述当前像素点在所述原始RGB图像中的像素值是否位于所述肤色的像素值范围内。
根据本发明实施例的一种图像处理系统，包括：
图像处理模块，用于将原始RGB图像进行双边滤波及边界滤波处理，以对应得到双边滤波图像及边界滤波图像；
像素值判断模块，用于当判断到所述原始RGB图像中的任意当前像素点为肤色的像素点时，判断所述当前像素点在所述边界滤波图像中的像素值是否小于阈值；
第一处理模块，用于判断到所述当前像素点在所述边界滤波图像中的像素值不小于所述阈值时，将所述当前像素点在所述原始RGB图像中的像素值作为所述当前像素点对应的美颜像素值；
第二处理模块，用于判断到所述当前像素点在所述边界滤波图像中的像素值小于所述阈值时，将所述当前像素点在所述原始RGB图像及所述双边滤波图像上的像素值进行混合运算，并将运算出的混合像素值作为所述当前像素点对应的所述美颜像素值；
美颜渲染模块，用于根据每个所述当前像素点对应的所述美颜像素值，将所述原始RGB图像进行美颜渲染。
另外，根据本发明上述实施例的一种图像处理系统，还可以具有如下附加的技术特征：
进一步地，所述图像处理系统还包括：
像素点判断模块，用于判断所述当前像素点是否为所述肤色的像素点；
当判断到所述当前像素点不为所述肤色的像素点时，所述第一处理模块将所述当前像素点在所述原始RGB图像中的像素值作为所述当前像素点对应的美颜像素值。
进一步地，所述图像处理系统还包括：
美白运算模块，用于将所述当前像素点对应的所述美颜像素值进行美白运算，以得到所述当前像素点对应的美白像素值；
美白渲染模块，用于根据每个所述当前像素点对应的所述美白像素值，将所述原始RGB图像进行美白渲染；
其中，所述将所述当前像素点对应的所述美颜像素值进行美白运算的公式为：
w(xi,yi)为所述当前像素点对应的所述美颜像素值，v(xi,yi)为所述当前像素点进行美白运算后得到的对应的所述美白像素值，β为亮度增强系数。
本发明还提出一种计算机可读存储介质，其上存储有计算机程序，该程序被处理器执行时实现上述的图像处理方法。
本发明还提出一种移动摄像设备，包括存储器、处理器以及存储在存储器上并可在处理器上运行的计算机程序，所述处理器为GPU，所述处理器执行所述程序时实现上述的图像处理方法。
上述图像处理方法、系统、可读存储介质及移动摄像设备，通过对所述原始RGB图像进行边界滤波处理，以得到所述边界滤波图像，然后对所述原始RGB图像中的每个像素点进行判断，当判断到当前像素点是肤色且其在所述边界滤波图像的像素值不小于所述阈值时，代表该像素点为肤色中的边界像素点，则在对该像素点进行美颜渲染时，采用该像素点在所述原始RGB图像中的原始像素值，当判断到当前像素点是肤色且其在所述边界滤波图像的像素值小于所述阈值时，代表该像素点为肤色且为非边界像素点，则在对该像素点进行美颜渲染时，采用该像素点在所述原始RGB图像及所述双边滤波图像中的像素值的
混合的综合像素值，因此所述图像处理方法、系统、可读存储介质及移动摄像设备在对肤色进行美颜的同时能够较好的保护眉毛、头发等边界细节特征，且对肤色且非边界像素点进行了综合美颜考虑，故美颜后的图像失真度小，更贴切于真实图片。
图1为本发明第一实施例中的图像处理方法的流程图。
图2为本发明第二实施例中的图像处理方法的流程图。
图3为本发明第三实施例中的图像处理系统的结构示意图。
主要元件符号说明：
| 图像处理模块 | 11 | 像素点判断模块 | 12 |
| 第一处理模块 | 13 | 第二处理模块 | 14 |
| 美颜渲染模块 | 15 | 肤色判断单元 | 121 |
| 阈值判断单元 | 122 | 美白运算模块 | 16 |
| 美白渲染模块 | 17 | 图像转换模块 | 18 |
| 肤色判断子单元 | 1211 |
以下具体实施方式将结合上述附图进一步说明本发明。
为了便于理解本发明，下面将参照相关附图对本发明进行更全面的描述。附图中给出了本发明的若干实施例。但是，本发明可以以许多不同的形式来实现，并不限于本文所描述的实施例。相反地，提供这些实施例的目的是使对本发明的公开内容更加透彻全面。
需要说明的是，当元件被称为“固设于”另一个元件，它可以直接在另一个元件上或者也可以存在居中的元件。当一个元件被认为是“连接”另一个元件，它可以是直接连接到另一个元件或者可能同时存在居中元件。本文所使用的术语“垂直的”、“水平的”、“左”、“右”以及类似的表述只是为了说明的目的。
除非另有定义，本文所使用的所有的技术和科学术语与属于本发明的技术领域的技术人员通常理解的含义相同。本文中在本发明的说明书中所使用的术语只是为了描述具体的实施例的目的，不是旨在于限制本发明。本文所使用的术语“及/或”包括一个或多个相关的所列项目的任意的和所有的组合。
请参阅图1，所示为本发明第一实施例中的图像处理方法，包括步骤S01至步骤S03。
步骤S01，将原始RGB图像进行双边滤波及边界滤波处理，以对应得到双边滤波图像及边界滤波图像。
其中，通过双边滤波对所述原始RGB图像进行处理的目的在于保边去噪，在实现对所述原始RGB图像进行处理的同时，保护了所述原始RGB图像的边缘。所述边界滤波处理也称Canny滤波处理，而Canny滤波处理是基于Canny算法来实现的，通过Canny滤波处理的目的在于寻找出所述原始RGB图像的边界。
步骤S02，当判断到所述原始RGB图像中的任意当前像素点为肤色的像素点时，判断所述当前像素点在所述边界滤波图像中的像素值是否小于阈值。
其中，所述步骤S02可以按照所述原始RGB图像中像素点的排列顺序来依序对每个像素点进行判断。
需要指出的是，肤色的像素值在一定范围内，通过判断所述原始RGB图像中的像素点的像素值是否在所述肤色(肤色像素值)范围内，即可判断出该像素点是否为肤色。
此外，还需要指出的是，当所述原始RGB图像中的像素点在所述边界滤波图像中的像素值不小于所述阈值时，代表该像素点在图像中为变化比较剧烈的部位，该像素点为边界，同理，当所述原始RGB图像中的像素点在所述边界滤波图像中的像素值小于所述阈值时，代表该像素点在图像中变化均匀，该像素点为非边界。
当所述步骤S02判断到所述当前像素点在所述边界滤波图像中的像素值不小于所述阈值，代表该像素点为肤色中的边界(如肤色与眉毛的交界处)像素点，则执行步骤S03，当所述步骤S02判断到所述当前像素点在所述边界滤波图
像中的像素值小于所述阈值时，代表该像素点为肤色且为非边界像素点，则执行步骤S04。
步骤S03，将所述当前像素点在所述原始RGB图像中的像素值作为所述当前像素点对应的美颜像素值。
可以理解的，所述步骤S03的目的在于，当判断到所述当前像素点为肤色中的边界像素点时，直接采用该像素点的原始像素值作为后续的美颜渲染，为此，较好的保护了肤色中的边界。
此外，还需要指出的是，当判断到所述当前像素点不为所述肤色的像素点时，则执行所述步骤S03，较好的保护了所述原始RGB图像中的周围环境(如背景)。
步骤S04，将所述当前像素点在所述原始RGB图像及所述双边滤波图像上的像素值进行混合运算，并将运算出的混合像素值作为所述当前像素点对应的所述美颜像素值。
其中，将所述当前像素点在所述原始RGB图像及所述双边滤波图像上的像素值进行混合运算的过程为：对所述当前像素点在所述原始RGB图像的像素值及在所述双边滤波图像上的像素值分配一定的权重比例，然后在进行混合运算，最终得到一个综合的像素值，即为所述美颜像素值。
故，所述步骤S04的目的在于，当判断到所述当前像素点为肤色且为非边界像素点时，通过将该像素点的原始像素值和双边滤波后的像素值进行综合，实现了在对肤色且非边界像素点进行美颜的同时，尽量使其贴切于真实状态，失真度小。
步骤S05，根据每个所述当前像素点对应的所述美颜像素值，将所述原始RGB图像进行美颜渲染。
可以理解的，通过所述步骤S03及所述步骤S04可以得到所述原始RGB图像当中每个像素点对应的所述美颜像素值，然后通过将每个所述美颜像素值渲染在对应的像素点上，即可完成对所述原始RGB图像的美颜。
综上，本发明上述实施例当中的图像处理方法，通过对所述原始RGB图像进行边界滤波处理，以得到所述边界滤波图像，然后对所述原始RGB图像中的
每个像素点进行判断，当判断到当前像素点不为肤色时，代表该像素点为周围环境的像素点，则在对该像素点进行美颜渲染时，采用该像素点在所述原始RGB图像中的原始像素值，当判断到当前像素点是肤色且其在所述边界滤波图像的像素值不小于所述阈值时，代表该像素点为肤色中的边界像素点，则在对该像素点进行美颜渲染时，采用该像素点在所述原始RGB图像中的原始像素值，当判断到当前像素点是肤色且其在所述边界滤波图像的像素值小于所述阈值时，代表该像素点为肤色且为非边界像素点，则在对该像素点进行美颜渲染时，采用该像素点在所述原始RGB图像及所述双边滤波图像中的像素值的混合的综合像素值，因此所述图像处理方法、系统、可读存储介质及移动摄像设备在对肤色进行美颜的同时能够较好的保护周围环境(如背景)、眉毛、头发等边界细节特征，且对肤色且非边界像素点进行了综合美颜考虑，故美颜后的图像失真度小，更贴切于真实图片。
请参阅图2，所示为本发明第二实施例中的图像处理方法，包括步骤S11至步骤S18。
步骤S11，将摄像头获取的YUV图像转换为原始RGB图像。
需要指出的是，摄像头获取的图像通常为YUV图像，固需要将获取的YUV图像转换为对应的原始RGB图像，以便于后续的美颜操作。
此外，可以采用现有的YUV与RGB的格式转换公式来完成，将摄像头获取的YUV图像转换为原始RGB图像。
步骤S12，将所述原始RGB图像进行双边滤波及边界滤波处理，以对应得到双边滤波图像及边界滤波图像。
其中，将所述原始RGB图像进行双边滤波处理的公式为：
其中，
WP用于归一化，Ifiltered(x)指滤波后的图像，I(x)指原始图像，xi指当前滤波处理的像素点的坐标，Ω指像
素点坐标中心点，gs是空间的卷积核，fr指值域核。
步骤S13，判断所述原始RGB图像中的任意当前像素点是否为肤色的像素点。
其中，所述判断所述当前像素点是否为所述肤色像素点的步骤包括：
判断所述当前像素点在所述原始RGB图像中的像素值是否位于所述肤色的像素值范围内。当判断到所述当前像素点在所述原始RGB图像中的像素值位于所述肤色的像素值范围内时，则判断出所述当前像素点为所述肤色的像素点，即为肤色像素点，当判断到所述当前像素点在所述原始RGB图像中的像素值不位于所述肤色的像素值范围内时，则判断出所述当前像素点不为所述肤色的像素点，即为非肤色像素点。
当所述步骤S13判断到所述当前像素点不为所述肤色的像素点时，代表该像素点为周围环境，则依次执行步骤S14、步骤S17、步骤S18及步骤S19，当所述步骤S14判断到所述当前像素点为所述肤色的像素点时，代表该像素点为肤色，则执行步骤S15。
步骤S14，将所述当前像素点在所述原始RGB图像中的像素值作为所述当前像素点对应的美颜像素值。
步骤S15，判断所述当前像素点在所述边界滤波图像中的像素值是否小于阈值。
当所述步骤S15判断到所述当前像素点在所述边界滤波图像中的像素值是小于所述阈值时，代表该像素点为肤色且为非边界像素点，则执行步骤S16至步骤S19，当所述步骤S15判断到所述当前像素点在所述边界滤波图像中的像素值不是小于所述阈值时，则代表该像素点为肤色中的边界像素点，则返回执行所述步骤S14。
步骤S16，将所述当前像素点在所述原始RGB图像及所述双边滤波图像上的像素值进行混合运算，并将运算出的混合像素值作为所述当前像素点对应的所述美颜像素值。
其中，将所述当前像素点在所述原始RGB图像及所述双边滤波图像上的像
素值进行混合运算的公式为：(1-smooth)*(A(xi,yi)-B(xi,yi))+B(xi,yi)，
其中，A(xi,yi)为所述当前像素点在所述原始RGB图像中的像素值，B(xi,yi)为所述当前像素点在所述双边滤波图像上的像素值，smooth为光滑度，可以根据实际情况进行调整，当光滑度越高，双边滤波后的图像所占比例越大，光滑度为1时，所述当前像素点的所述美颜像素值直接采用所述当前像素点在所述双边滤波图像上的像素值，当光滑度为0时，所述当前像素点的所述美颜像素值直接采用其在所述原始RGB图像上的原始像素值。
步骤S17，根据每个所述当前像素点对应的所述美颜像素值，将所述原始RGB图像进行美颜渲染。
步骤S18，将所述当前像素点对应的所述美颜像素值进行美白运算，以得到所述当前像素点对应的美白像素值。
其中，所述将所述当前像素点对应的所述美颜像素值进行美白运算的公式为：
w(xi,yi)为所述当前像素点对应的所述美颜像素值，v(xi,yi)为所述当前像素点进行美白运算后得到的对应的所述美白像素值，β为亮度增强系数，可以根据实际情况进行调节。
步骤S19，根据每个所述当前像素点对应的所述美白像素值，将所述原始RGB图像进行美白渲染。
需要指出的是，所述步骤S18及所述步骤S19，利用log曲线将原图的色阶映射为两端稍弱，中间稍强，可以将画面亮度自然的增强，达到美白的效果。
综上，本发明上述实施例当中的图像处理方法，相较于第一实施里当中的图像处理方法，其在完成对所述原始RGB图像进行美颜之后，还将对美颜之后的整个图像进行美白处理，以进一步地提高美颜之后的图像效果。
本发明另一方面还提供一种图像处理系统，请查阅图3，所示为本发明第三实施例中的图像处理系统，所述图像处理系统为GPU，所述图像处理系统设置于移动摄像设备上(如手机、平板等)，包括：
图像处理模块11，用于将原始RGB图像进行双边滤波及边界滤波处理，以
对应得到双边滤波图像及边界滤波图像；
像素值判断模块12，用于当判断到所述原始RGB图像中的任意当前像素点为肤色的像素点时，判断所述当前像素点在所述边界滤波图像中的像素值是否小于阈值；
第一处理模块13，用于判断到所述当前像素点在所述边界滤波图像中的像素值不小于所述阈值时，将所述当前像素点在所述原始RGB图像中的像素值作为所述当前像素点对应的美颜像素值；
第二处理模块14，用于判断到所述当前像素点在所述边界滤波图像中的像素值小于所述阈值时，将所述当前像素点在所述原始RGB图像及所述双边滤波图像上的像素值进行混合运算，并将运算出的混合像素值作为所述当前像素点对应的所述美颜像素值；
美颜渲染模块15，用于根据每个所述当前像素点对应的所述美颜像素值，将所述原始RGB图像进行美颜渲染。
其中，将所述原始RGB图像进行双边滤波处理的公式为：
其中，
WP用于归一化，Ifiltered(x)指滤波后的图像，I(x)指原始图像，xi指当前滤波处理的像素点的坐标，Ω指像素点坐标中心点，gs是空间的卷积核，fr指值域核；
将所述当前像素点在所述原始RGB图像及所述双边滤波图像上的像素值进行混合运算的公式为：(1-smooth)*(A(xi,yi)-B(xi,yi))+B(xi,yi)，
其中，smooth为光滑度，A(xi,yi)为所述当前像素点在所述原始RGB图像中的像素值，B(xi,yi)为所述当前像素点在所述双边滤波图像上的像素值。
进一步地，所述图像处理系统还包括：
像素点判断模块16，用于判断所述当前像素点是否为所述肤色的像素点；
当判断到所述当前像素点不为所述肤色的像素点时，所述第一处理模块13
将所述当前像素点在所述原始RGB图像中的像素值作为所述当前像素点对应的美颜像素值。
进一步地，所述图像处理系统还包括：
美白运算模块17，用于将所述当前像素点对应的所述美颜像素值进行美白运算，以得到所述当前像素点对应的美白像素值；
美白渲染模块18，用于根据每个所述当前像素点对应的所述美白像素值，将所述原始RGB图像进行美白渲染；
其中，所述将所述当前像素点对应的所述美颜像素值进行美白运算的公式为：
w(xi,yi)为所述当前像素点对应的所述美颜像素值，v(xi,yi)为所述当前像素点进行美白运算后得到的对应的所述美白像素值，β为亮度增强系数。
进一步地，所述图像处理系统还包括：
图像转换模块19，用于将摄像头获取的YUV图像转换为所述原始RGB图像。
进一步地，所述像素点判断模块16包括：
像素点判断子单元161，用于判断所述原始RGB图像中的每个所述像素点的像素值是否在肤色范围内。
在本实施例当中，由于摄像头获取的图像直接由GPU处理，并且整个图像处理及美白的过程均在GPU上完成，相比于现有技术，可以较好的释放CPU的运算资源，同时也加速整个图像处理过程，从而在移动平台上也可以实现实时美颜的效果。
本发明还提出一种计算机可读存储介质，其上存储有计算机程序，该程序被处理器执行时实现如上述的图像处理方法。
本发明还提出一种移动摄像设备，包括存储器、处理器以及存储在存储器上并可在处理器上运行的计算机程序，所述处理器为GPU，所述处理器执行所述程序时实现如上述的图像处理方法。
本领域技术人员可以理解，在流程图中表示或在此以其他方式描述的逻辑和/或步骤，例如，可以被认为是用于实现逻辑功能的可执行指令的定序列表，可以具体实现在任何计算机可读介质中，以供指令执行系统、装置或设备(如基于计算机的系统、包括处理器的系统或其他可以从指令执行系统、装置或设备取指令并执行指令的系统)使用，或结合这些指令执行系统、装置或设备而使用。就本说明书而言，“计算机可读介质”可以是任何可以包含、存储、通信、传播或传输程序以供指令执行系统、装置或设备或结合这些指令执行系统、装置或设备而使用的装置。
计算机可读介质的更具体的示例(非穷尽性列表)包括以下：具有一个或多个布线的电连接部(电子装置)，便携式计算机盘盒(磁装置)，随机存取存储器(RAM)，只读存储器(ROM)，可擦除可编辑只读存储器(EPROM或闪速存储器)，光纤装置，以及便携式光盘只读存储器(CDROM)。另外，计算机可读介质甚至可以是可在其上打印所述程序的纸或其他合适的介质，因为可以例如通过对纸或其他介质进行光学扫描，接着进行编辑、解译或必要时以其他合适方式进行处理来以电子方式获得所述程序，然后将其存储在计算机存储器中。
应当理解，本发明的各部分可以用硬件、软件、固件或它们的组合来实现。在上述实施方式中，多个步骤或方法可以用存储在存储器中且由合适的指令执行系统执行的软件或固件来实现。例如，如果用硬件来实现，和在另一实施方式中一样，可用本领域公知的下列技术中的任一项或它们的组合来实现：具有用于对数据信号实现逻辑功能的逻辑门电路的离散逻辑电路，具有合适的组合逻辑门电路的专用集成电路，可编程门阵列(PGA)，现场可编程门阵列(FPGA)等。
在本说明书的描述中，参考术语“一个实施例”、“一些实施例”、“示例”、“具体示例”、或“一些示例”等的描述意指结合该实施例或示例描述的具体特征、结构、材料或者特点包含于本发明的至少一个实施例或示例中。在本说明书中，对上述术语的示意性表述不一定指的是相同的实施例或示例。而且，描述的具体特征、结构、材料或者特点可以在任何的一个或多个实施例或示例中以合适
的方式结合。
以上所述实施例仅表达了本发明的几种实施方式，其描述较为具体和详细，但并不能因此而理解为对本发明专利范围的限制。应当指出的是，对于本领域的普通技术人员来说，在不脱离本发明构思的前提下，还可以做出若干变形和改进，这些都属于本发明的保护范围。因此，本发明专利的保护范围应以所附权利要求为准。
Claims (10)
- 一种图像处理方法，其特征在于，包括：将原始RGB图像进行双边滤波及边界滤波处理，以对应得到双边滤波图像及边界滤波图像；当判断到所述原始RGB图像中的任意当前像素点为肤色的像素点时，判断所述当前像素点在所述边界滤波图像中的像素值是否小于阈值；若否，则将所述当前像素点在所述原始RGB图像中的像素值作为所述当前像素点对应的美颜像素值；若是，则将所述当前像素点在所述原始RGB图像及所述双边滤波图像上的像素值进行混合运算，并将运算出的混合像素值作为所述当前像素点对应的所述美颜像素值；根据每个所述当前像素点对应的所述美颜像素值，将所述原始RGB图像进行美颜渲染。
- 根据权利要求1所述的图像处理方法，其特征在于，在所述将原始RGB图像进行双边滤波及边界滤波处理，以对应得到双边滤波图像及边界滤波图像的步骤之后，还包括：判断所述当前像素点是否为所述肤色的像素点；若否，则执行所述将所述当前像素点在所述原始RGB图像中的像素值作为所述当前像素点对应的美颜像素值的步骤。
- 根据权利要求1所述的图像处理方法，其特征在于，将所述原始RGB图像进行双边滤波处理的公式为：
其中，WP用于归一化，Ifiltered(x)指滤波后的图像，I(x)指原始图像，xi指当前滤波处理的像素点的坐标，Ω指像素点坐标中心点，gs是空间的卷积核，fr指值域核；
将所述当前像素点在所述原始RGB图像及所述双边滤波图像上的像素值进行混合运算的公式为：(1-smooth)*(A(xi,yi)-B(xi,yi))+B(xi,yi)，其中，smooth为光滑度，A(xi,yi)为所述当前像素点在所述原始RGB图像中的像素值，B(xi,yi)为所述当前像素点在所述双边滤波图像上的像素值。 - 根据权利要求1所述的图像处理方法，其特征在于，在所述根据每个所述当前像素点对应的所述美颜像素值，将所述原始RGB图像进行美颜渲染的步骤之后，还包括：将所述当前像素点对应的所述美颜像素值进行美白运算，以得到所述当前像素点对应的美白像素值；根据每个所述当前像素点对应的所述美白像素值，将所述原始RGB图像进行美白渲染；其中，所述将所述当前像素点对应的所述美颜像素值进行美白运算的公式为：w(xi,yi)为所述当前像素点对应的所述美颜像素值，v(xi,yi)为所述当前像素点进行美白运算后得到的对应的所述美白像素值，β为亮度增强系数。
- 根据权利要求2所述的图像处理方法，其特征在于，所述判断所述当前像素点是否为所述肤色的像素点的步骤包括：判断所述当前像素点在所述原始RGB图像中的像素值是否位于所述肤色的像素值范围内。
- 一种图像处理系统，其特征在于，包括：图像处理模块，用于将原始RGB图像进行双边滤波及边界滤波处理，以对应得到双边滤波图像及边界滤波图像；像素值判断模块，用于当判断到所述原始RGB图像中的任意当前像素点为肤色的像素点时，判断所述当前像素点在所述边界滤波图像中的像素值是否小于阈值；第一处理模块，用于判断到所述当前像素点在所述边界滤波图像中的像素值不小于所述阈值时，将所述当前像素点在所述原始RGB图像中的像素值作为所述当前像素点对应的美颜像素值；第二处理模块，用于判断到所述当前像素点在所述边界滤波图像中的像素值小于所述阈值时，将所述当前像素点在所述原始RGB图像及所述双边滤波图像上的像素值进行混合运算，并将运算出的混合像素值作为所述当前像素点对应的所述美颜像素值；美颜渲染模块，用于根据每个所述当前像素点对应的所述美颜像素值，将所述原始RGB图像进行美颜渲染。
- 根据权利要求6所述的图像处理系统，其特征在于，所述图像处理系统还包括：像素点判断模块，用于判断所述当前像素点是否为所述肤色的像素点；当判断到所述当前像素点不为所述肤色的像素点时，所述第一处理模块将所述当前像素点在所述原始RGB图像中的像素值作为所述当前像素点对应的美颜像素值。
- 根据权利要求6所述的图像处理系统，其特征在于，所述图像处理系统还包括：美白运算模块，用于将所述当前像素点对应的所述美颜像素值进行美白运算，以得到所述当前像素点对应的美白像素值；美白渲染模块，用于根据每个所述当前像素点对应的所述美白像素值，将所述原始RGB图像进行美白渲染；其中，所述将所述当前像素点对应的所述美颜像素值进行美白运算的公式为：w(xi,yi)为所述当前像素点对应的所述美颜像素值，v(xi,yi)为所述当前像素点进行美白运算后得到的对应的所述美白像素值，β为亮度增强系数。
- 一种计算机可读存储介质，其上存储有计算机程序，其特征在于，该程序被处理器执行时实现如权利要求1－5任一所述的方法。
- 一种移动摄像设备，包括存储器、处理器以及存储在存储器上并可在处理器上运行的计算机程序，其特征在于，所述处理器为GPU，所述处理器执行 所述程序时实现如权利要求1－5任一所述的方法。
Applications Claiming Priority (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710805304.9 | 2017-09-08 | ||
| CN201710805304.9A CN107808404A (zh) | 2017-09-08 | 2017-09-08 | 图像处理方法、系统、可读存储介质及移动摄像设备 |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| WO2019047409A1 true WO2019047409A1 (zh) | 2019-03-14 |
Family
ID=61576562
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| PCT/CN2017/116362 WO2019047409A1 (zh) | 2017-09-08 | 2017-12-15 | 图像处理方法、系统、可读存储介质及移动摄像设备 |
Country Status (2)
| Country | Link |
|---|---|
| CN (1) | CN107808404A (zh) |
| WO (1) | WO2019047409A1 (zh) |
Families Citing this family (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109639982B (zh) * | 2019-01-04 | 2020-06-30 | Oppo广东移动通信有限公司 | 一种图像降噪方法、装置、存储介质及终端 |
| CN110363717B (zh) * | 2019-06-28 | 2021-07-23 | 北京字节跳动网络技术有限公司 | 一种处理面部图像的方法、装置、介质和电子设备 |
| CN111462003B (zh) * | 2020-03-20 | 2022-08-23 | 稿定（厦门）科技有限公司 | 人脸图像处理方法、介质、设备及装置 |
| CN111669611B (zh) * | 2020-06-19 | 2022-02-22 | 广州繁星互娱信息科技有限公司 | 图像处理方法、装置、终端及存储介质 |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20090185757A1 (en) * | 2008-01-22 | 2009-07-23 | Samsung Electronics Co., Ltd. | Apparatus and method for immersion generation |
| CN101930601A (zh) * | 2010-09-01 | 2010-12-29 | 浙江大学 | 一种基于边缘信息的多尺度模糊图像盲复原方法 |
| CN105787888A (zh) * | 2014-12-23 | 2016-07-20 | 联芯科技有限公司 | 人脸图像美化方法 |
| CN105874506A (zh) * | 2013-10-04 | 2016-08-17 | 谷歌公司 | 具有细节保留的图像模糊 |
| CN106550243A (zh) * | 2016-12-09 | 2017-03-29 | 武汉斗鱼网络科技有限公司 | 直播视频处理方法、装置及电子设备 |
Family Cites Families (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN100544400C (zh) * | 2007-08-27 | 2009-09-23 | 北京航空航天大学 | 结合可见光图像信息的sar图像斑点噪声抑制方法 |
| CN106920211A (zh) * | 2017-03-09 | 2017-07-04 | 广州四三九九信息科技有限公司 | 美颜处理方法、装置以及终端设备 |
-
2017
- 2017-09-08 CN CN201710805304.9A patent/CN107808404A/zh active Pending
- 2017-12-15 WO PCT/CN2017/116362 patent/WO2019047409A1/zh active Application Filing
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20090185757A1 (en) * | 2008-01-22 | 2009-07-23 | Samsung Electronics Co., Ltd. | Apparatus and method for immersion generation |
| CN101930601A (zh) * | 2010-09-01 | 2010-12-29 | 浙江大学 | 一种基于边缘信息的多尺度模糊图像盲复原方法 |
| CN105874506A (zh) * | 2013-10-04 | 2016-08-17 | 谷歌公司 | 具有细节保留的图像模糊 |
| CN105787888A (zh) * | 2014-12-23 | 2016-07-20 | 联芯科技有限公司 | 人脸图像美化方法 |
| CN106550243A (zh) * | 2016-12-09 | 2017-03-29 | 武汉斗鱼网络科技有限公司 | 直播视频处理方法、装置及电子设备 |
Also Published As
| Publication number | Publication date |
|---|---|
| CN107808404A (zh) | 2018-03-16 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US9639956B2 (en) | Image adjustment using texture mask | |
| US9311901B2 (en) | Variable blend width compositing | |
| WO2018188534A1 (zh) | 人脸图像处理方法、装置和电子设备 | |
| WO2016206087A1 (zh) | 一种低照度图像处理方法和装置 | |
| WO2019047409A1 (zh) | 图像处理方法、系统、可读存储介质及移动摄像设备 | |
| WO2023123927A1 (zh) | 图像增强方法、装置、设备和存储介质 | |
| KR102317613B1 (ko) | 로컬라이징된 콘트라스트 향상을 위한 시스템들 및 방법들 | |
| US9881202B2 (en) | Providing visual effects for images | |
| US20090220169A1 (en) | Image enhancement | |
| US8731248B2 (en) | Method of performing eye circle correction an image and related computing device | |
| US8879835B2 (en) | Fast adaptive edge-aware matting | |
| CN105243371A (zh) | 一种人脸美颜程度的检测方法、系统及拍摄终端 | |
| CN109859217B (zh) | 人脸图像中毛孔区域的分割方法及计算设备 | |
| WO2014169579A1 (zh) | 一种色彩增强方法及装置 | |
| CN109919859B (zh) | 一种户外场景图像去雾增强方法、计算设备及其存储介质 | |
| JP6841888B2 (ja) | 映像処理装置及び映像処理方法 | |
| US20130121565A1 (en) | Method and Apparatus for Local Region Selection | |
| WO2019090580A1 (en) | System and method for image dynamic range adjusting | |
| CN111145086A (zh) | 一种图像处理方法、装置及电子设备 | |
| WO2015103739A1 (zh) | 增强图像对比度的装置、电子设备及方法 | |
| CN112819691B (zh) | 图像处理方法、装置、设备及可读存储介质 | |
| CN114581979A (zh) | 图像处理方法和装置 | |
| WO2017161542A1 (en) | Skin map-aided skin smoothing of images using a bilateral filter | |
| CN107564085B (zh) | 图像扭曲处理方法、装置、计算设备及计算机存储介质 | |
| CN109636749B (zh) | 图像处理方法 |
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: 17924560 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: 17924560 Country of ref document: EP Kind code of ref document: A1 |
|
| 32PN | Ep: public notification in the ep bulletin as address of the adressee cannot be established |
Free format text: NOTING OF LOSS OF RIGHTS PURSUANT TO RULE 112(1) EPC (EPO FORM 1205A DATED 16.10.2020) |
|
| 122 | Ep: pct application non-entry in european phase |
Ref document number: 17924560 Country of ref document: EP Kind code of ref document: A1 |