CN113793247A - Ornament image beautifying method and terminal - Google Patents
Ornament image beautifying method and terminal Download PDFInfo
- Publication number
- CN113793247A CN113793247A CN202110770880.0A CN202110770880A CN113793247A CN 113793247 A CN113793247 A CN 113793247A CN 202110770880 A CN202110770880 A CN 202110770880A CN 113793247 A CN113793247 A CN 113793247A
- Authority
- CN
- China
- Prior art keywords
- image
- jewelry
- information
- data
- calculation formula
- Prior art date
- Legal status (The legal status 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 status listed.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 28
- 238000012545 processing Methods 0.000 claims abstract description 32
- 230000011218 segmentation Effects 0.000 claims abstract description 27
- 238000004364 calculation method Methods 0.000 claims description 43
- 238000001914 filtration Methods 0.000 claims description 29
- 238000004422 calculation algorithm Methods 0.000 claims description 22
- 230000004927 fusion Effects 0.000 claims description 10
- 238000004590 computer program Methods 0.000 claims description 8
- 230000014759 maintenance of location Effects 0.000 claims description 6
- 230000009466 transformation Effects 0.000 claims description 6
- 230000008859 change Effects 0.000 claims description 5
- 238000009795 derivation Methods 0.000 claims description 5
- 230000000694 effects Effects 0.000 abstract description 9
- 238000010276 construction Methods 0.000 abstract description 5
- 238000003384 imaging method Methods 0.000 abstract description 5
- 238000012805 post-processing Methods 0.000 description 5
- 230000008569 process Effects 0.000 description 5
- 230000009286 beneficial effect Effects 0.000 description 3
- 230000003993 interaction Effects 0.000 description 2
- 239000002184 metal Substances 0.000 description 2
- 230000035807 sensation Effects 0.000 description 2
- 238000005282 brightening Methods 0.000 description 1
- 239000003086 colorant Substances 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 230000007613 environmental effect Effects 0.000 description 1
- 230000006870 function Effects 0.000 description 1
- 238000003709 image segmentation Methods 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 239000000463 material Substances 0.000 description 1
- 230000009467 reduction Effects 0.000 description 1
- 210000000697 sensory organ Anatomy 0.000 description 1
- 230000000007 visual effect Effects 0.000 description 1
Images
Classifications
-
- G06T3/04—
Abstract
The invention discloses a jewelry image beautifying method and a terminal, wherein an initial image is obtained, and is identified and segmented to obtain jewelry image information and jewelry category data; performing structure identification and structure segmentation on jewelry image information according to jewelry category data to obtain each part image and part outline hierarchical data; beautification processing is carried out on each part image, and after beautification processing, each part image is fused according to the part outline layered data to form a complete jewelry image; performing super-resolution processing on the jewelry image to obtain a final image; the invention realizes the automation of jewelry image beautification, reduces the cost required by the post-beautification treatment of the jewelry image, reduces the post-treatment difficulty of the jewelry image, improves the jewelry shooting imaging efficiency, and avoids the problems of untimely construction period and incapability of meeting the real-time effect viewing requirement because the post-treatment difficulty of the jewelry image is reduced and other parties are not required to carry out treatment.
Description
Technical Field
The invention relates to the technical field of image beautification, in particular to a jewelry image beautification method and a jewelry image beautification terminal.
Background
The jewelry image display plays an important role in the commercial activity of jewelry sales, and particularly the display of product effects on a product album and an e-commerce website is directly related to the most visual first sense organ of a buyer to the jewelry. Therefore, the merchant can be fully developed in jewelry shooting and image processing, and aims to produce high-quality jewelry pictures. However, the jewelry has a complex structure, and the shooting difficulty is increased due to the problems of mirror reflection, darkness, uneven light and the like during shooting, so that professional post-processing is required, a merchant needs to spend a lot of money for obtaining a high-quality jewelry image, and the image interaction is often not timely and low in efficiency.
And the existing jewelry image post-production is manually edited and beautified by utilizing an image modifying tool, the workload is large, and the beautifying steps cannot be unified and standardized. And because the manual work is finished, high labor cost is needed, and the problems that the image beautification interaction period is not timely and the real-time effect viewing requirement cannot be met are also caused.
Disclosure of Invention
The technical problem to be solved by the invention is as follows: the jewelry image beautifying method and the terminal can automatically beautify the jewelry image, reduce the cost and meet the requirements of users.
In order to solve the technical problems, the invention adopts the technical scheme that:
a jewelry image beautification method, comprising:
s1, acquiring an initial image, and identifying and segmenting the initial image to obtain jewelry image information and jewelry category data;
s2, performing structure identification and structure segmentation on the jewelry image information according to the jewelry category data to obtain each part image and part outline hierarchical data;
s3, performing beautification treatment on each part image, and fusing each part image according to the part contour hierarchical data after the beautification treatment to form a complete jewelry image;
and S4, performing super-resolution processing on the jewelry image to obtain a final image.
In order to solve the technical problem, the invention adopts another technical scheme as follows:
a jewelry image beautification terminal comprises a processor, a memory and a computer program which is stored on the memory and can run on the processor, wherein the processor executes the computer program to realize the following steps:
s1, acquiring an initial image, and identifying and segmenting the initial image to obtain jewelry image information and jewelry category data;
s2, performing structure identification and structure segmentation on the jewelry image information according to the jewelry category data to obtain each part image and part outline hierarchical data;
s3, performing beautification treatment on each part image, and fusing each part image according to the part contour hierarchical data after the beautification treatment to form a complete jewelry image;
and S4, performing super-resolution processing on the jewelry image to obtain a final image.
The invention has the beneficial effects that: the invention realizes the automation of jewelry image beautification, reduces the cost required by the jewelry image post-beautification treatment, reduces the difficulty of jewelry image post-treatment, improves the jewelry shooting imaging efficiency, and avoids the problems of untimely construction period and incapability of meeting the real-time effect viewing requirement because the difficulty of jewelry image post-treatment can be reduced, a merchant can carry out treatment by himself without entrusting other parties for treatment.
Drawings
FIG. 1 is a flow chart of a method for beautifying jewelry image according to an embodiment of the present invention;
FIG. 2 is a block diagram of a jewelry image beautification terminal according to an embodiment of the present invention;
FIG. 3 is a detailed flowchart of a jewelry image beautification method according to an embodiment of the present invention;
description of reference numerals:
1. a jewelry image beautifying terminal; 2. a processor; 3. a memory.
Detailed Description
In order to explain technical contents, achieved objects, and effects of the present invention in detail, the following description is made with reference to the accompanying drawings in combination with the embodiments.
Referring to fig. 1 and fig. 3, a jewelry image beautifying method includes:
s1, acquiring an initial image, and identifying and segmenting the initial image to obtain jewelry image information and jewelry category data;
s2, performing structure identification and structure segmentation on the jewelry image information according to the jewelry category data to obtain each part image and part outline hierarchical data;
s3, performing beautification treatment on each part image, and fusing each part image according to the part contour hierarchical data after the beautification treatment to form a complete jewelry image;
and S4, performing super-resolution processing on the jewelry image to obtain a final image.
From the above description, the beneficial effects of the present invention are: the invention realizes the automation of jewelry image beautification, reduces the cost required by the jewelry image post-beautification treatment, reduces the difficulty of jewelry image post-treatment, improves the jewelry shooting imaging efficiency, and avoids the problems of untimely construction period and incapability of meeting the real-time effect viewing requirement because the difficulty of jewelry image post-treatment can be reduced, a merchant can carry out treatment by himself without entrusting other parties for treatment.
Further, the step S3 is specifically:
s31, carrying out edge and detail protection filtering processing on each component image to obtain a first image;
s32, performing standard color fitting processing on the first image to obtain a second image;
s33, carrying out curve adjustment of the image layer on the second image, and carrying out image layer fusion with the original image layer through a high contrast retention algorithm to obtain a third image;
and S34, fusing the third images of the components.
According to the description, the method applies the edge-preserving detail-preserving filtering algorithm, the standard color fitting, the image processing, the image fusion and other methods, and quickly solves the problems of mirror reflection, darkness, uneven light sensation and the like generated in the jewelry shooting process.
Further, the step S1 is specifically:
s11, acquiring an initial image, loading a semantic segmentation program, and performing image semantic segmentation and identification on the jewelry to obtain jewelry information, background information, jewelry category data and jewelry edge contour data;
s12, segmenting an initial image according to the jewelry information, the background information and the jewelry edge contour data to obtain jewelry image information, wherein the jewelry image information comprises the jewelry edge contour data;
the step S2 specifically includes:
s21, performing semantic segmentation and identification on the jewelry image information according to the jewelry category information and the jewelry edge contour data to obtain each part information and corresponding part contour data of the jewelry;
and S22, segmenting the jewelry image information according to the part information and the part outline data to obtain part images of all parts and corresponding part outline hierarchical data.
As can be seen from the above description, the jewelry and the background in the image are segmented according to the contour, and each part of the jewelry is further segmented, so that the specific processing is respectively carried out, and meanwhile, part contour hierarchical data is generated for the fusion of the subsequent part images.
Further, the step S31 includes:
filtering each part image by adopting rapid edge-preserving filtering, wherein in a local mean value calculation formula of the rapid edge-preserving filtering, a current pixel xijThe integral image in the domain range with width n and height m is represented as the sum of all pixels from the upper left coordinate (i-n, j-m), which is again denoted as (k, l), to the lower right coordinate (i + n, j + m), where the local mean m of the current pixel isijThe calculation formula of (a) is as follows:
its local mean square error vi，jThe calculation formula of (2) is as follows:
after the local mean and the local mean square error are obtained through calculation, new pixel values
The filter calculation formula is:
wherein:
wherein, sigma is a constant parameter;
the formula for calculating the variance is obtained by carrying out mathematical derivation according to the relation between the variance and the mean value as follows:
from the above description, the edge preserving and detail preserving filtering processing is performed on each segmented part image, wherein one filtering method is rapid edge preserving filtering, and the edge preserving fuzzy algorithm is an edge preserving fuzzy algorithm for realizing local mean square error through an integral image, and the calculation is simple and the calculated amount is independent of the radius.
Further, the step S32 is specifically:
starting a color picking program, obtaining standard color values corresponding to each part image through transformation and comparison of color values in a color space, and fitting the standard color values into the image through an Alpha value algorithm for saturation and brightness, wherein a color value space change formula is as follows:
R′＝R/255
G′＝G/255
B′＝B/255
Cmax＝max(R′，G′，B′)
Cmin＝min(R′，G′，B′)
Δ＝Cmax-Cmin
the chromaticity calculation formula is as follows:
the saturation calculation formula:
the brightness calculation formula is as follows:
as can be seen from the above description, the present application makes each part present a relatively ideal color by obtaining the standard color values corresponding to each part image, and simultaneously fitting the standard color values into the image through an Alpha value algorithm for saturation and brightness.
Referring to fig. 2, a jewelry image beautifying terminal includes a processor, a memory and a computer program stored on the memory and capable of running on the processor, wherein the processor executes the computer program to implement the following steps:
s1, acquiring an initial image, and identifying and segmenting the initial image to obtain jewelry image information and jewelry category data;
s2, performing structure identification and structure segmentation on the jewelry image information according to the jewelry category data to obtain each part image and part outline hierarchical data;
s3, performing beautification treatment on each part image, and fusing each part image according to the part contour hierarchical data after the beautification treatment to form a complete jewelry image;
and S4, performing super-resolution processing on the jewelry image to obtain a final image.
From the above description, the beneficial effects of the present invention are: the invention realizes the automation of jewelry image beautification, reduces the cost required by the jewelry image post-beautification treatment, reduces the difficulty of jewelry image post-treatment, improves the jewelry shooting imaging efficiency, and avoids the problems of untimely construction period and incapability of meeting the real-time effect viewing requirement because the difficulty of jewelry image post-treatment can be reduced, a merchant can carry out treatment by himself without entrusting other parties for treatment.
Further, the step S3 is specifically:
s31, carrying out edge and detail protection filtering processing on each component image to obtain a first image;
s32, performing standard color fitting processing on the first image to obtain a second image;
s33, carrying out curve adjustment of the image layer on the second image, and carrying out image layer fusion with the original image layer through a high contrast retention algorithm to obtain a third image;
and S34, fusing the third images of the components.
According to the description, the method applies the edge-preserving detail-preserving filtering algorithm, the standard color fitting, the image processing, the image fusion and other methods, and quickly solves the problems of mirror reflection, darkness, uneven light sensation and the like generated in the jewelry shooting process.
Further, the step S1 is specifically:
s11, acquiring an initial image, loading a semantic segmentation program, and performing image semantic segmentation and identification on the jewelry to obtain jewelry information, background information, jewelry category data and jewelry edge contour data;
s12, segmenting an initial image according to the jewelry information, the background information and the jewelry edge contour data to obtain jewelry image information, wherein the jewelry image information comprises the jewelry edge contour data;
the step S2 specifically includes:
s21, performing semantic segmentation and identification on the jewelry image information according to the jewelry category information and the jewelry edge contour data to obtain each part information and corresponding part contour data of the jewelry;
and S22, segmenting the jewelry image information according to the part information and the part outline data to obtain part images of all parts and corresponding part outline hierarchical data.
As can be seen from the above description, the jewelry and the background in the image are segmented according to the contour, and each part of the jewelry is further segmented, so that the specific processing is respectively carried out, and meanwhile, part contour hierarchical data is generated for the fusion of the subsequent part images.
Further, the step S31 includes:
filtering each part image by adopting rapid edge-preserving filtering, wherein in a local mean value calculation formula of the rapid edge-preserving filtering, a current pixel xijThe integral image in the domain range with width n and height m is represented as the sum of all pixels from the upper left coordinate (i-n, j-m), which is again denoted as (k, l), to the lower right coordinate (i + n, j + m), where the local mean m of the current pixel isijThe calculation formula of (a) is as follows:
its local mean square error vi，jThe calculation formula of (2) is as follows:
after the local mean and the local mean square error are obtained through calculation, new pixel values
The filter calculation formula is:
wherein:
wherein, sigma is a constant parameter;
the formula for calculating the variance is obtained by carrying out mathematical derivation according to the relation between the variance and the mean value as follows:
from the above description, the edge preserving and detail preserving filtering processing is performed on each segmented part image, wherein one filtering method is rapid edge preserving filtering, and the edge preserving fuzzy algorithm is an edge preserving fuzzy algorithm for realizing local mean square error through an integral image, and the calculation is simple and the calculated amount is independent of the radius.
Further, the step S32 is specifically:
starting a color picking program, obtaining standard color values corresponding to each part image through transformation and comparison of color values in a color space, and fitting the standard color values into the image through an Alpha value algorithm for saturation and brightness, wherein a color value space change formula is as follows:
R′＝R/255
G′＝G/255
B′＝B/255
Cmax＝max(R′，G′，B′)
Cmin＝min(R′，G′，B′)
Δ＝Cmax-Cmin
the chromaticity calculation formula is as follows:
the saturation calculation formula:
the brightness calculation formula is as follows:
as can be seen from the above description, the present application makes each part present a relatively ideal color by obtaining the standard color values corresponding to each part image, and simultaneously fitting the standard color values into the image through an Alpha value algorithm for saturation and brightness.
Referring to fig. 1 and fig. 3, a first embodiment of the present invention is:
a jewelry image beautification method, comprising:
s1, acquiring an initial image, and identifying and segmenting the initial image to obtain jewelry image information and jewelry category data;
the step S1 specifically includes:
s11, acquiring an initial image, loading a semantic segmentation program, and performing image semantic segmentation and identification on the jewelry to obtain jewelry information, background information, jewelry category data and jewelry edge contour data;
s12, segmenting the initial image according to the jewelry information, the background information and the jewelry edge contour data to obtain jewelry image information, wherein the jewelry image information comprises the jewelry edge contour data.
In the embodiment, environmental factors such as lighting of a studio are adjusted during shooting, shooting background paper and jewelry are placed in the studio, a shooting distance and jewelry shooting mobile application are adjusted, a shooting button is clicked in an interface, an application shooting function interface is entered, a scene to be shot in the studio is presented in a shooting preview interface, and meanwhile, a camera image monitors or obtains image data of each frame, so that an initial image is obtained, and an image beautification engine is created. After the image beautification engine is started, a semantic segmentation program is loaded at first, the jewelry is subjected to first image semantic segmentation and recognition, background information and main body information in the image are recognized at the moment, and main body objects, namely jewelry edge contour data of the jewelry and jewelry information are obtained.
After the jewelry edge contour data is obtained, the initial image is segmented according to the jewelry edge contour data, and an image of a jewelry main body is extracted to obtain jewelry image information.
S2, performing structure identification and structure segmentation on the jewelry image information according to the jewelry category data to obtain each part image and part outline hierarchical data;
the step S2 specifically includes:
s21, performing semantic segmentation and identification on the jewelry image information according to the jewelry category information and the jewelry edge contour data to obtain each part information and corresponding part contour data of the jewelry;
and S22, segmenting the jewelry image information according to the part information and the part outline data to obtain part images of all parts and corresponding part outline hierarchical data.
In the embodiment in the market, after jewelry image information is obtained, a semantic segmentation program is sent again to perform image segmentation with higher granularity for the second time, and the segmentation recognition is intended to extract information of each part in a picture main body. For example, in this example, identifying and segmenting an image of a ring will result in jewelry and ring part image information, as well as corresponding part profile hierarchy data.
S3, performing beautification treatment on each part image, and fusing each part image according to the part contour hierarchical data after the beautification treatment to form a complete jewelry image;
the step S3 specifically includes:
s31, carrying out edge and detail protection filtering processing on each component image to obtain a first image;
in this embodiment, after obtaining the component image information of each component, the image beautification engine will construct a set of filters to protect the edges of each component of the jewelry respectivelyAnd (5) carrying out detail-keeping filtering processing. In this embodiment, one of the adopted filtering is fast edge preserving filtering, which is an edge preserving fuzzy algorithm for realizing local mean square error through an integral image, and the calculation is simple and the calculated amount is independent of the radius. In the local mean calculation formula of the fast edge-preserving filter, the current pixel xijThe integral image in the domain range with width n and height m is represented as the sum of all pixels from the upper left coordinate (i-n, j-m), which is again denoted as (k, l), to the lower right coordinate (i + n, j + m), where the local mean m of the current pixel isijThe calculation formula of (a) is as follows:
its local mean square error vijThe calculation formula of (2) is as follows:
after the local mean and the local mean square error are obtained through calculation, new pixel values
The filter calculation formula is:
wherein:
wherein, sigma is a constant parameter;
the formula for calculating the variance is obtained by carrying out mathematical derivation according to the relation between the variance and the mean value as follows:
s32, performing standard color fitting processing on the first image to obtain a second image;
the step S32 specifically includes:
starting a color picking program, obtaining standard color values corresponding to each part image through transformation and comparison of color values in a color space, and fitting the standard color values into the image through an Alpha value algorithm for saturation and brightness, wherein a color value space change formula is as follows:
R′＝R/255
G′＝G/255
B′＝B/255
Cmax＝max(R′，G′，B′)
Cmin＝min(R′，G′，B′)
Δ＝Cmax-Cmin
the chromaticity calculation formula is as follows:
the saturation calculation formula:
the brightness calculation formula is as follows:
in this embodiment, the colors of the component images after the filtering are relatively balanced, and color fitting needs to be performed on the filtered images at this time. The color picking program is started firstly, the standard color value of the part is obtained through the transformation and comparison of color space color values, and the standard color value is fitted into the image through an Alpha value algorithm for saturation and brightness.
And S33, performing graph layer curve adjustment on the second image, and performing graph layer fusion with the original graph layer through a high contrast retention algorithm to obtain a third image.
In this embodiment, after color fitting is performed, the brightness and darkness details of each image are obtained by deepening and brightening the curve adjustment of the image layer. The newly-built brightness enhancement or darkness enhancement curve image layer aims at modifying the brightness of an image and cannot cause texture loss, the brightest channel and the darkest channel are loaded in the image in a reverse mode, the image obtains a proper contrast, and finally, the image layer fusion is carried out through a high contrast retention algorithm and the original image layer, so that the minimum texture is stored.
And S34, fusing the third images of the components.
In the embodiment, after the processing, the images of the parts basically restore the problems of reflection of light on the metal surface, darkness caused by insufficient light, brightness inconsistency and the like in the shooting process.
And S4, performing super-resolution processing on the jewelry image to obtain a final image.
In this embodiment, the image beautification engine will start the image super-resolution processing program to perform high-definition processing on the image, and finally beautify the jewelry image, so as to obtain a final image and display the final image in the preview interface.
Referring to fig. 2, the second embodiment of the present invention is:
a jewelry image beautifying terminal 1 comprises a processor 2, a memory 3 and a computer program which is stored on the memory 3 and can run on the processor 2, wherein the processor 2 realizes the steps in the first embodiment when executing the computer program.
In conclusion, the jewelry image beautifying method and the terminal provided by the invention realize the automation of jewelry image beautifying, reduce the cost required by the post-beautifying processing of the jewelry image, reduce the post-processing difficulty of the jewelry image, and improve the jewelry shooting imaging efficiency, and due to the reduction of the post-processing difficulty of the jewelry image, a merchant can process the jewelry image by himself without entrusting other parties for processing, thereby avoiding the problems that the construction period is not timely and the real-time effect viewing requirement cannot be met; meanwhile, the method utilizes a program algorithm mode to beautify various problems caused by metal reflection, uneven light and the like during shooting, aims to solve the jewelry shooting problem in a fast and standardized mode, improves the image post-processing efficiency, not only can beautify and process images of jewelry in real time during jewelry shooting, but also can play the same role in post-processing of shot images in different fields by collecting characteristic materials and training and modifying algorithm parameters.
The above description is only an embodiment of the present invention, and not intended to limit the scope of the present invention, and all equivalent changes made by using the contents of the present specification and the drawings, or applied directly or indirectly to the related technical fields, are included in the scope of the present invention.
Claims (10)
1. A jewelry image beautifying method is characterized by comprising the following steps:
s1, acquiring an initial image, and identifying and segmenting the initial image to obtain jewelry image information and jewelry category data;
s2, performing structure identification and structure segmentation on the jewelry image information according to the jewelry category data to obtain each part image and part outline hierarchical data;
s3, performing beautification treatment on each part image, and fusing each part image according to the part contour hierarchical data after the beautification treatment to form a complete jewelry image;
and S4, performing super-resolution processing on the jewelry image to obtain a final image.
2. The jewelry image beautifying method according to claim 1, wherein the step S3 is specifically:
s31, carrying out edge and detail protection filtering processing on each component image to obtain a first image;
s32, performing standard color fitting processing on the first image to obtain a second image;
s33, carrying out curve adjustment of the image layer on the second image, and carrying out image layer fusion with the original image layer through a high contrast retention algorithm to obtain a third image;
and S34, fusing the third images of the components.
3. A jewelry image beautifying method according to claim 2, wherein the step S1 specifically is:
s11, acquiring an initial image, loading a semantic segmentation program, and performing image semantic segmentation and identification on the jewelry to obtain jewelry information, background information, jewelry category data and jewelry edge contour data;
s12, segmenting an initial image according to the jewelry information, the background information and the jewelry edge contour data to obtain jewelry image information, wherein the jewelry image information comprises the jewelry edge contour data;
the step S2 specifically includes:
s21, performing semantic segmentation and identification on the jewelry image information according to the jewelry category information and the jewelry edge contour data to obtain each part information and corresponding part contour data of the jewelry;
and S22, segmenting the jewelry image information according to the part information and the part outline data to obtain part images of all parts and corresponding part outline hierarchical data.
4. A jewelry image beautifying method according to claim 2, wherein said step S31 comprises:
filtering each part image by adopting rapid edge-preserving filtering, wherein in a local mean value calculation formula of the rapid edge-preserving filtering, a current pixel xijThe integral image in the domain range with width n and height m is represented as the sum of all pixels from the upper left coordinate (i-n, j-m), which is again denoted as (k, l), to the lower right coordinate (i + n, j + m), where the local mean m of the current pixel isijThe calculation formula of (a) is as follows:
its local mean square error vijThe calculation formula of (2) is as follows:
after the local mean and the local mean square error are obtained through calculation, new pixel values
The filter calculation formula is:
wherein:
wherein, sigma is a constant parameter;
the formula for calculating the variance is obtained by carrying out mathematical derivation according to the relation between the variance and the mean value as follows:
5. a jewelry image beautifying method according to claim 2, wherein the step S32 specifically is:
starting a color picking program, obtaining standard color values corresponding to each part image through transformation and comparison of color values in a color space, and fitting the standard color values into the image through an Alpha value algorithm for saturation and brightness, wherein a color value space change formula is as follows:
R'＝R/155
G'＝G/155
B′＝B/255
Cmax＝max(R′，G′，B′)
Cmin＝min(R′，G′，B′)
Δ＝Cmax-Cmin
the chromaticity calculation formula is as follows:
the saturation calculation formula:
the brightness calculation formula is as follows:
6. a jewelry image beautification terminal comprises a processor, a memory and a computer program which is stored on the memory and can run on the processor, and is characterized in that the processor executes the computer program to realize the following steps:
s1, acquiring an initial image, and identifying and segmenting the initial image to obtain jewelry image information and jewelry category data;
s2, performing structure identification and structure segmentation on the jewelry image information according to the jewelry category data to obtain each part image and part outline hierarchical data;
s3, performing beautification treatment on each part image, and fusing each part image according to the part contour hierarchical data after the beautification treatment to form a complete jewelry image;
and S4, performing super-resolution processing on the gesture image model to obtain a final image.
7. A jewelry image beautifying terminal according to claim 6, wherein the step S3 is specifically:
s31, carrying out edge and detail protection filtering processing on each component image to obtain a first image;
s32, performing standard color fitting processing on the first image to obtain a second image;
s33, carrying out curve adjustment of the image layer on the second image, and carrying out image layer fusion with the original image layer through a high contrast retention algorithm to obtain a third image;
and S34, fusing the third images of the components.
8. A jewelry image beautifying terminal according to claim 7, wherein the step S1 is specifically:
s11, acquiring an initial image, loading a semantic segmentation program, and performing image semantic segmentation and identification on the jewelry to obtain jewelry information, background information, jewelry category data and jewelry edge contour data;
s12, segmenting an initial image according to the jewelry information, the background information and the jewelry edge contour data to obtain jewelry image information, wherein the jewelry image information comprises the jewelry edge contour data;
the step S2 specifically includes:
s21, performing semantic segmentation and identification on the jewelry image information according to the jewelry category information and the jewelry edge contour data to obtain each part information and corresponding part contour data of the jewelry;
and S22, segmenting the jewelry image information according to the part information and the part outline data to obtain part images of all parts and corresponding part outline hierarchical data.
9. A jewelry image beautifying terminal according to claim 7, wherein said step S31 comprises:
filtering each part image by adopting rapid edge-preserving filtering, wherein in a local mean value calculation formula of the rapid edge-preserving filtering, a current pixel xijThe integral image in the domain range with width n and height m is represented as the sum of all pixels from the upper left coordinate (i-n, j-m), which is again denoted as (k, l), to the lower right coordinate (i + n, j + m), where the local mean m of the current pixel isijThe calculation formula of (a) is as follows:
its local mean square error vijThe calculation formula of (2) is as follows:
after the local mean and the local mean square error are obtained through calculation, new pixel values
The filter calculation formula is:
wherein:
wherein, sigma is a constant parameter;
the formula for calculating the variance is obtained by carrying out mathematical derivation according to the relation between the variance and the mean value as follows:
10. a jewelry image beautifying terminal according to claim 7, wherein the step S32 is specifically:
starting a color picking program, obtaining standard color values corresponding to each part image through transformation and comparison of color values in a color space, and fitting the standard color values into the image through an Alpha value algorithm for saturation and brightness, wherein a color value space change formula is as follows:
R′＝R/255
G′＝G/255
B′＝B/255
Cmax＝max(R′，G′，B′)
Cmin＝min(R′，G′，B′)
Δ＝Cmax-Cmin
the chromaticity calculation formula is as follows:
the saturation calculation formula:
the brightness calculation formula is as follows:
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110770880.0A CN113793247A (en) | 2021-07-08 | 2021-07-08 | Ornament image beautifying method and terminal |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110770880.0A CN113793247A (en) | 2021-07-08 | 2021-07-08 | Ornament image beautifying method and terminal |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN113793247A true CN113793247A (en) | 2021-12-14 |
Family
ID=79181052
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110770880.0A Pending CN113793247A (en) | 2021-07-08 | 2021-07-08 | Ornament image beautifying method and terminal |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113793247A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115423827A (en) * | 2022-11-03 | 2022-12-02 | 北京百度网讯科技有限公司 | Image processing method, image processing device, electronic equipment and storage medium |
-
2021
- 2021-07-08 CN CN202110770880.0A patent/CN113793247A/en active Pending
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115423827A (en) * | 2022-11-03 | 2022-12-02 | 北京百度网讯科技有限公司 | Image processing method, image processing device, electronic equipment and storage medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Gharbi et al. | Deep bilateral learning for real-time image enhancement | |
| Chen et al. | Robust image and video dehazing with visual artifact suppression via gradient residual minimization | |
| Bonneel et al. | Example-based video color grading. | |
| US7187788B2 (en) | Method and system for enhancing portrait images that are processed in a batch mode | |
| Pouli et al. | Progressive color transfer for images of arbitrary dynamic range | |
| TWI704524B (en) | Method and device for image polishing | |
| Winnemöller | Xdog: advanced image stylization with extended difference-of-gaussians | |
| AU2003204466B2 (en) | Method and system for enhancing portrait images | |
| CN102938837B (en) | A kind of tone mapping method keeping full sub-model based on edge | |
| CN111105359A (en) | Tone mapping method for high dynamic range image | |
| Liu et al. | Progressive complex illumination image appearance transfer based on CNN | |
| CN109961016A (en) | The accurate dividing method of more gestures towards Intelligent household scene | |
| Song et al. | Multi-scale joint network based on Retinex theory for low-light enhancement | |
| CN113793247A (en) | Ornament image beautifying method and terminal | |
| Liba et al. | Sky optimization: Semantically aware image processing of skies in low-light photography | |
| DiPaola | Painterly rendered portraits from photographs using a knowledge-based approach | |
| CN114155569B (en) | Cosmetic progress detection method, device, equipment and storage medium | |
| Zhao et al. | Learning tone curves for local image enhancement | |
| Gao et al. | PencilArt: a chromatic penciling style generation framework | |
| CN110264564B (en) | Film and television style simulation method based on color chart mapping and histogram mapping | |
| US20210150677A1 (en) | Systems and Methods for Selective Enhancement of Skin Features in Images | |
| WO2019023957A1 (en) | Image capturing method and image capturing system of intelligent terminal | |
| Pavan Kumar et al. | A refined structure preserving image abstraction framework as a pre-processing technique for desire focusing on prominent structure and artistic stylization | |
| Zavalishin et al. | Visually aesthetic image contrast enhancement | |
| CN111915500A (en) | Foggy day image enhancement method based on improved Retinex algorithm |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |