WO2021128731A1 - 图像处理方法及装置、图像处理设备及存储介质 - Google Patents

图像处理方法及装置、图像处理设备及存储介质 Download PDF

Info

Publication number
WO2021128731A1
WO2021128731A1 PCT/CN2020/093442 CN2020093442W WO2021128731A1 WO 2021128731 A1 WO2021128731 A1 WO 2021128731A1 CN 2020093442 W CN2020093442 W CN 2020093442W WO 2021128731 A1 WO2021128731 A1 WO 2021128731A1
Authority
WO
WIPO (PCT)
Prior art keywords
points
grid
area
point
key
Prior art date
Application number
PCT/CN2020/093442
Other languages
English (en)
French (fr)
Inventor
李通
刘文韬
钱晨
Original Assignee
北京市商汤科技开发有限公司
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 北京市商汤科技开发有限公司 filed Critical 北京市商汤科技开发有限公司
Priority to KR1020207037116A priority Critical patent/KR20210084348A/ko
Priority to SG11202109179WA priority patent/SG11202109179WA/en
Priority to JP2020572416A priority patent/JP7160958B2/ja
Publication of WO2021128731A1 publication Critical patent/WO2021128731A1/zh
Priority to US17/377,444 priority patent/US20210342970A1/en

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T3/00—Geometric image transformations in the plane of the image
    • G06T3/04—Context-preserving transformations, e.g. by using an importance map
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/70—Determining position or orientation of objects or cameras
    • G06T7/73—Determining position or orientation of objects or cameras using feature-based methods
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T3/00—Geometric image transformations in the plane of the image
    • G06T3/18—Image warping, e.g. rearranging pixels individually
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/10—Segmentation; Edge detection
    • G06T7/11—Region-based segmentation
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/70—Determining position or orientation of objects or cameras
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V10/00—Arrangements for image or video recognition or understanding
    • G06V10/20—Image preprocessing
    • G06V10/25—Determination of region of interest [ROI] or a volume of interest [VOI]
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
    • G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
    • G06V40/103—Static body considered as a whole, e.g. static pedestrian or occupant recognition
    • 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
    • G06T2207/30—Subject of image; Context of image processing
    • G06T2207/30196—Human being; Person

Definitions

  • the present disclosure relates to the field of image technology, and in particular to an image processing method and device, image processing equipment and storage medium.
  • the embodiments of the present disclosure expect to provide an image processing method and device, image processing equipment, and storage medium.
  • a first aspect of the embodiments of the present disclosure provides an image processing method, including: acquiring position information of a first key point of a first part contained in a target object in a first image; and determining whether to include The first area of the first key point; the first area is determined according to the relative position between the grid points of the deformed grid preset in the first area and the pixel points in the first area The displacement of the inner grid points; according to the displacement of the grid points in the first area, the deformation of the pixel points in the first area is controlled to obtain a deformed second image.
  • the first area is determined according to the relative position between the grid points of the deformed grid and the pixel points in the first area.
  • the displacement of the grid points in a region includes: determining the grid points according to the relative positions between the grid points of the preset deformed grid in the first region and the pixel points in the first region The attenuation parameter of the displacement of the grid point; and, according to the deformation instruction, the first displacement of the grid point is determined; and the attenuation process is performed on the first displacement according to the attenuation parameter to obtain a ratio of the first displacement The second displacement is small.
  • the controlling the deformation of the pixel points in the first area according to the displacement of the grid points in the first area to obtain the deformed second image includes: The second displacement amount controls the distance between adjacent pixels in the first area to obtain a deformed second image.
  • the grid point is determined according to the relative position between the grid point of the deformed grid preset in the first area and the pixel in the first area
  • the attenuation parameter of the displacement amount includes: determining the pixel points located on the line of the plurality of the first key points to obtain the first set; according to the pixel points in the first set and the grid in the deformed grid The relative positions of the points to obtain a second set, where the second set includes: the target grid points in the deformed grid that are closest to each pixel in the first set; The relative position between each of the target grid points and the pixel points in the first set controlled by the target grid points to determine the attenuation of each of the target grid points in the second set parameter.
  • each of the target grid points in the second set and the pixel points in the first set controlled by the target grid points Position includes: traversing each of the second set in the second set outward in a predetermined direction with each of the first key points as the center.
  • the target grid points are sorted by the distances of the target grid points in the second set relative to each of the first key points in a predetermined direction; according to the distance sorting, the second The attenuation parameter of each of the target grid points in the set.
  • the determining the attenuation parameter of each target grid point in the second set according to the spacing order includes: any one in the second set When the target grid point is located in a predetermined direction of the plurality of first key points, determine the candidate value of the attenuation parameter according to the sorting of the distances corresponding to the plurality of first key points; select The maximum value among the candidate values is used as the attenuation parameter of any one of the target grid points.
  • the method further includes: determining the second region corresponding to the second part according to the position information of the second key point of the second part contained in the target in the first image; According to the first displacement amount of the grid points of the preset deformed grid in the second area, the deformation of the second area in the first image is controlled to obtain the deformed second image.
  • An image processing device disclosed in an embodiment of the present disclosure includes: an acquisition module configured to acquire position information of a first key point of a first part contained in a target object in a first image; a first determination module configured to be based on the The location information of the first key point determines the first area containing the first key point; the second determining module is configured to compare the grid points of the deformed grid preset in the first area with the first area. The relative position between the pixel points in the area determines the displacement of the grid points in the first area; the control module is configured to control the first area according to the displacement of the grid points in the first area The inner pixel is deformed to obtain the deformed second image.
  • the second determining module is configured to determine the distance between the grid points of the preset deformed grid in the first area and the pixel points in the first area Determine the attenuation parameter of the displacement of the grid point; and, according to the deformation instruction, determine the first displacement of the grid point; perform attenuation processing on the first displacement according to the attenuation parameter , To obtain a second displacement smaller than the first displacement.
  • control module is configured to control the distance between adjacent pixels in the first area according to the second displacement amount to obtain a deformed second image .
  • the second determining module is configured to determine the pixel points located on the line of the plurality of first key points to obtain the first set; The relative positions of the pixel points and the grid points in the deformed grid to obtain a second set, where the second set includes: the target in the deformed grid that is closest to each pixel in the first set Grid points; according to the relative position between each of the target grid points in the second set and the pixel points in the first set controlled by the target grid points, determine the second set The attenuation parameter of each target grid point.
  • the second determining module is configured to respectively take each of the first key points as the center and traverse each of the targets in the second set outward in a predetermined direction. Grid points, obtain the spacing order of each target grid point in the second set relative to each of the first key points in a predetermined direction; according to the spacing order, determine each target grid point in the second set The attenuation parameter of the target grid point.
  • the second determining module is configured to: when any one of the target grid points in the second set is located in a predetermined direction of a plurality of the first key points , Determining the candidate value of the attenuation parameter according to the sorting of the distances corresponding to the plurality of the first key points; selecting the maximum value among the candidate values as the all of the target grid points The attenuation parameters.
  • the first part is an upper limb;
  • the acquisition module is configured to acquire position information of a skeleton key point of the upper limb in the first image, and the skeleton key point includes the following key At least one of the points: shoulder key points, elbow joint key points, wrist key points, and hand key points.
  • the first determining module is further configured to determine the second key point corresponding to the second part according to the position information of the second key point of the second part contained in the target in the first image Second area
  • the control module is further configured to control the deformation of the second area in the first image according to the first displacement of the grid points of the preset deformed grid in the second area to obtain the deformed first Two images.
  • a third aspect of the embodiments of the present disclosure provides an image processing device, including: a memory; a processor connected to the memory and configured to implement the image provided by any of the foregoing technical solutions by executing computer executable instructions stored on the memory Approach.
  • a fourth aspect of the embodiments of the present disclosure provides a computer storage medium that stores computer-executable instructions; after the computer-executable instructions are executed by a processor, the image processing method provided by any of the foregoing technical solutions can be implemented.
  • the first key point of the first part before deforming the entire first image using the deformed grid, the first key point of the first part can be determined first, and then the first area that needs to be protected can be obtained based on the first key point
  • the displacement of the grid points in the first area is determined according to the relative position between the grid points and the pixels in the first area, and is no longer just based on a single Deformation instructions to determine.
  • the relative position between the grid points and the pixel points is introduced to determine the displacement of the grid points, so that precise control of the displacement of the grid points in the first area can be realized, and the difference in the same image can be realized. Fine control of the deformation of the pixels in the area, thereby effectively improving the image deformation effect.
  • FIG. 1 is a schematic flowchart of an image processing method provided by an embodiment of the disclosure
  • 2A is a usage diagram of a standard deformed grid laid on a first image provided by an embodiment of the disclosure
  • 2B is a schematic diagram of a first area and a second area provided by an embodiment of the present disclosure
  • FIG. 3 is a schematic diagram of a process for determining attenuation parameters according to an embodiment of the disclosure
  • FIG. 4 is a schematic diagram of a connection of key points and a second set of grid points provided by an embodiment of the present disclosure
  • FIG. 5 is a schematic structural diagram of an image processing device provided by an embodiment of the disclosure.
  • FIG. 6 is a schematic structural diagram of an image processing device provided by an embodiment of the disclosure.
  • first, second, third, etc. may be used to describe various information in the embodiments of the present disclosure, the information should not be limited to these terms. These terms are only used to distinguish the same type of information from each other.
  • first information may also be referred to as second information, and similarly, the second information may also be referred to as first information.
  • word “if” as used herein can be interpreted as "when” or “when” or “in response to a certainty”.
  • an embodiment of the present disclosure provides an image processing method, including:
  • Step S110 Obtain the position information of the first key point of the first part contained in the target object in the first image
  • Step S120 Based on the location information of the first key point, determine the first area containing the first key point;
  • Step S130 Determine the displacement of the grid points in the first area according to the relative positions between the grid points of the deformed grid preset in the first area and the pixel points in the first area;
  • Step S140 According to the displacement of the grid points in the first area, control the deformation of the pixel points in the first area to obtain a deformed second image.
  • the image processing method provided in this embodiment can be applied to electronic devices with image processing functions.
  • the image device may include various terminal devices, and the terminal device includes: a mobile phone or a wearable device.
  • the terminal device may also include: a vehicle-mounted terminal device, or a fixed terminal device dedicated to image collection and fixed in a certain place.
  • the image device may further include a server, for example, a local server or a cloud server located in a cloud platform that provides image processing services.
  • the target object is, for example, a human body, an animal, or a virtual object rendered from a virtual three-dimensional model, and the present disclosure does not limit the specific form of the target object.
  • the first part of the target object may be a limb part.
  • the first part of the target object may be an arm, a leg, or an abdomen, which is not limited in the embodiment of the present disclosure.
  • the first image before performing image deformation processing, the first image may be divided into multiple regions, and the first region may include one or more of the multiple regions.
  • the first region may be a region containing the first part that needs to be deformed, or the first region may be a region that contains the first part that needs to be deformed.
  • the area of the first part that needs to be deformed may be an area where the degree of deformation is relatively large; the area of the first part that needs to be restrained of deformation may be an area that needs to be deformed relatively small.
  • the deformed grid may be determined. For example, laying a deformed grid on the first image.
  • the deformed grid may include: grid points formed by the intersection of horizontal and vertical lines.
  • the horizontal lines contained in the deformed grid can be referred to as latitude lines, and the vertical lines can be referred to as warp lines; the lines in the deformed grid can be collectively referred to as latitude and longitude lines.
  • the warp and weft lines may be straight lines respectively laid in the standard horizontal and vertical directions.
  • the deformation amplitude of each area in the image is the same, which may make the area that does not need to be deformed, the degree of deformation is small, or the area that needs to be deformed is larger. It will be deformed indiscriminately according to the uniform deformation amplitude, and this deformation method according to the uniform deformation amplitude may cause inconsistencies in the generated second image, resulting in poor deformation effect of the first image.
  • Figure 2A is a schematic diagram of a pre-deformed grid laid out.
  • Fig. 2B takes the portrait of Fig. 2A as an example, and the right upper limb area is determined as the first area.
  • the first key point of the first part contained in the target in the first image is first determined.
  • the first key point may be a skeleton key point or a contour key point of the first part
  • the skeleton key point is a key point where a human bone or an animal bone is located.
  • the contour key points may be the key points of the contour presented on the surface of the human body or animal. It can be understood that the first key point may be a point located on the first part, and may be used to locate the point of the first part, and the distribution position of the skeleton key point determines the position of the first part in the first image. Therefore, in this embodiment, the first region may be determined based on the location of one or more skeleton key points.
  • step S120 at least the boundary of the first area may be determined based on the position information of the first key point; the first area may be determined based on the boundary of the first area.
  • step S130 after the first area is determined, pixels in the first area can be obtained.
  • the deformation method of using a preset deformation grid to perform image deformation may be referred to as grid deformation.
  • the pixels in the first area are the pixels that may need to be moved. After the positions of the pixels are moved, the spacing between the pixels will change. Changes.
  • S140 may include: adjusting the density of pixels in the first area according to the displacement of the grid points in the first area, so as to control the deformation of the pixels in the first area to obtain the deformation After the second image.
  • the key points of the first part in the first image are determined, and then the pixel points in the first area are obtained based on the first key points; when the image is deformed, the displacement of the grid points in the first area is based on the The relative position between the grid points and the pixel points in the area is determined, and is no longer only determined based on a single deformation instruction.
  • precise control of the displacement of the grid points in the first area can be used to achieve fine control of pixel deformation (that is, the spacing between pixel points) in different areas of the image, thereby enhancing the effect of image deformation.
  • S130 may include: determining the grid points of the grid points according to the relative positions between the grid points of the preset deformed grid in the first area and the pixel points in the first area. The attenuation parameter of the displacement; and, according to the deformation command, the first displacement of the grid point is determined; the first displacement is attenuated according to the attenuation parameter to obtain a second displacement smaller than the first displacement.
  • the distance between a specific pixel in the first area and each grid point can be obtained according to the relative position between the grid point and the pixel point in the first area, and then the grid point can be determined according to the distance.
  • the magnitude of the attenuation parameter of the displacement amount may be a pixel point at the location of the first key point, or the specific pixel point may be a pixel point near the location of the first key point. This is only an example of determining the attenuation parameter of the displacement of the grid points in the first region based on the above-mentioned relative position, and it is specifically not limited to this.
  • the deformation instruction may be an instruction generated based on user input received by a human-computer interaction interface, or a deformation instruction generated based on image preset processing functions such as one-key beauty or body beauty. For example, if there is an automatic waist thinning function for a portrait in an image, the image processing device will generate a corresponding deformation instruction according to the automatic waist thinning function.
  • the deformation instruction may carry a deformation parameter.
  • the deformation parameter may include the above-mentioned first displacement amount.
  • the attenuation parameter and the first displacement are used as known quantities to perform attenuation processing for each first displacement, so as to obtain a second displacement smaller than the first displacement.
  • the attenuation parameter is a parameter used to reduce the displacement of the grid points in the first region.
  • the deformation amplitude of the pixels in the first area is positively correlated with the displacement of the grid points in the first area, that is, the greater the displacement of the grid points, the greater the deformation amplitude of the pixels in the first area; correspondingly , The smaller the displacement of the grid points, the smaller the deformation amplitude of the pixel points in the first area.
  • the attenuation parameter includes but is not limited to at least one of the following:
  • the attenuation coefficient can also be called the attenuation ratio; for example, the original first displacement of each grid point in the first area is calculated according to the deformation command, and the first displacement is multiplied by the attenuation coefficient to obtain the first displacement. The final second displacement of each grid point in a region.
  • Attenuation value If the attenuation value is positive, the original first displacement can be subtracted from the attenuation value to obtain a second displacement that is smaller than the original first displacement.
  • step S140 may include: controlling the distance between adjacent pixels in the first area according to the second displacement amount to obtain a deformed second image.
  • the deformed first region is transformed from the feature of equal spacing between adjacent pixels to the feature of unequal spacing between adjacent pixels. For example, if the second displacement of the grid point A is greater than the second displacement of the grid point B, the change in the spacing between the pixels controlled by the grid point A is greater than that of the pixels controlled by the grid point B The amount of change in the spacing between.
  • the effect of a thin waist is achieved. If the distance between adjacent pixels contained in the area where the breast is located is increased, the effect of breast enhancement can be achieved under the condition that the total number of pixels corresponding to the breast remains unchanged.
  • the first displacement obtained based on the deformation command is reduced to the second displacement.
  • the displacement of the first area can be suppressed (that is, weakened).
  • the deformation amplitude makes the deformation amplitudes of the first region and other regions different, and meets the deformation requirements of different deformation amplitudes in different regions, thereby improving the deformation effect of the first image being transformed into the second image.
  • the grid is determined according to the relative position between the grid points of the preset deformed grid in the first area and the pixel points in the first area.
  • the attenuation parameter of the displacement of the grid point may include the following steps:
  • Step 301 Determine the pixel points located on the connecting line of the plurality of first key points to obtain the first set
  • Step 302 Obtain a second set according to the relative positions of the pixel points in the first set and the grid points in the deformed grid, where the second set includes: the target in the deformed grid that is closest to each pixel in the first set Grid point
  • Step 303 Determine the attenuation parameter of each target grid point in the second set according to the relative position between each target grid point in the second set and the pixel point controlled by the target grid point in the first set.
  • the adjacent first key points are directly connected to obtain one or more lines, and the pixels located on these lines form the first set.
  • the pixels on the line of the first key point constitute the first set.
  • the first set is formed by pixel points on the line of the first key point; the second set contains target grid points adjacent to the line. Therefore, the first set is a set of pixel points, and the second set is a set of grid points, specifically a set of target grid points.
  • the deformation of the first image is performed based on the grid points of the deformed grid.
  • the area enclosed by the target grid points in the second set is the aforementioned first area.
  • the pixel point controlled by the target grid point in the first set may be the pixel point closest to the target grid point. It can be understood that the attenuation parameter of each target grid point in the second set can be determined according to the relative position of the pixel point on the line formed by each target grid point and the first key point.
  • some of the target grid points in the second set are close to the first key point, and some are far away from the first key point.
  • the target grid points close to the first key point in the second set have a larger attenuation parameter than the target grid points far away from the first key point.
  • the pixels in the first set may be part of the pixels in the first area. In the remaining pixels outside the first set in the first area, some pixels are close to the pixels in the first set, and some pixels are far away from the pixels in the first set.
  • the pixels far away from the pixels in the first set are generally controlled by the target grid points far from the first key point. Therefore, the attenuation amplitude of the pixel points far away from the first key point is smaller than that of the pixels close to the first key point. Deformation amplitude.
  • step 303 may include: traversing each target grid point in the second set outwards in a predetermined direction with each first key point as the center, to obtain each target grid point in the second set.
  • the target grid points are sorted in a predetermined direction with respect to the spacing of each first key point; and the attenuation parameter of each target grid point in the second set is determined according to the spacing sort.
  • the attenuation parameter of each target grid point in the second set may be determined according to the minimum distance between each target grid point in the second set and each pixel point in the first set.
  • the above minimum distance is an optional example of the spacing used in the above spacing sorting. For example, suppose there are M pixel points in the first set, and N target grid points in the second set. Both M and N are positive integers.
  • the N target grid points have M distances from the M pixel points, and the minimum distance among the M distances corresponding to each target grid point in the N target grid points is determined.
  • the minimum distances corresponding to the N target grid points are sorted to obtain the above-mentioned pitch sorting.
  • the relative position between the grid point and the first key point can be characterized by the aforementioned spacing or spacing order.
  • the attenuation parameter of the grid point closest to the center of the first part may be smaller, which leads to the phenomenon that the image effect is not as expected.
  • each target grid point in the second set is relative to the first key point.
  • Sort the pitch in a predetermined direction may be: a direction in which deformation of the first part needs to be suppressed, or a direction in which deformation is prohibited.
  • the predetermined direction may be a direction with a larger angle between the horizontal line direction and the vertical line direction of the deformed grid and the line corresponding to the first key point.
  • the predetermined direction may be a direction with a larger angle between the horizontal line direction and the vertical line direction of the deformed grid and the overall extension direction of the line formed by the first key point.
  • the predetermined direction may be a direction that has a larger angle with the extending direction of the first part among the horizontal line direction and the vertical line direction of the deformed grid. Under normal circumstances, a grid point in the second set may only be located in a predetermined direction of a pixel point.
  • the aforementioned spacing ordering may have a certain correlation with the corresponding attenuation parameter. For example, if the attenuation parameter is directly used for the attenuation of the first displacement of the target grid point, the higher the spacing order, the smaller the value of the attenuation parameter.
  • the foregoing determining the attenuation parameters of the target grid points in the second set according to the spacing sorting includes: any target grid point in the second set is located in a predetermined direction of the plurality of first key points In this case, the candidate value of the attenuation parameter is determined according to the distance ranking corresponding to each first key point; the maximum value among the candidate values is selected as the attenuation parameter of any target grid point.
  • a target grid point in the second set is located in a predetermined direction of the plurality of first key points, a plurality of pitch rankings will be determined.
  • one pitch order corresponds to one attenuation parameter.
  • a target grid point corresponding to the distance ranking of different first key points can be obtained, and then multiple candidate parameter values of attenuation parameters can be obtained.
  • the maximum value among the candidate values of the multiple attenuation parameters is selected as the attenuation parameter of the target grid point, so as to ensure that the closer the line formed by the first key point to the target grid point, the greater the attenuation parameter Case.
  • the first part is the upper limb.
  • the upper limb may include at least one of an upper arm, a forearm, and/or a hand.
  • S110 may include: acquiring position information of the skeleton key points of the upper limb in the first image, the skeleton key points including at least one of the following key points: shoulder key points, elbow joint key Points, wrist points and hand points.
  • the connection of the above-mentioned first key point may be: a connection of at least one key point in sequence from the key point of the shoulder, the key point of the elbow joint, the key point of the wrist to the key point of the hand.
  • the pixels included in the first set may include at least one of the following: pixels on the line between the shoulder key point and the elbow joint key point, and the pixel point on the line between the elbow joint key point and the wrist key point. , The pixel point on the line between the key point of the wrist and the key point of the hand.
  • the method further includes: determining the second area corresponding to the second part according to the position information of the second key point of the second part contained in the target in the first image; The first displacement amount of the grid points of the preset deformed grid within controls the deformation of the second region in the first image to obtain the deformed second image.
  • the first part is different from the first part. Referring to FIG. 2B, the waist area other than the right upper limb can also be determined as the second area.
  • the first displacement amount of the grid points in the first area and the second area may both be the initial displacement amount determined according to the deformation instruction.
  • the deformation amplitude can be controlled by the displacement of the grid points. Therefore, in this embodiment, the first area may be an area that needs to be deformed, and the second area may be an area that needs to be deformed. In the deformation process using the preset deformed grid, the deformation amplitude of the first region can be made smaller than the deformation amplitude of the second region based on the attenuation parameter and based on the same deformation instruction.
  • the deformation direction corresponding to the deformation amplitude includes but is not limited to at least one of the following: increase, decrease, rotation, mirror image, change of line shape, etc. of the deformation part of the corresponding region.
  • the first part of the deformation processing is the waist
  • the waist reduction processing using the preset deformation grid when the waist is compressed toward the center of the portrait, it is located near the waist
  • the arm may be stretched and deformed.
  • the image area where the arm is located can be set as the first area, and the image area where the waist is located is the second area.
  • the deformation amplitude of the first area is small, while the deformation amplitude of the second area is large; so, on the one hand
  • the effect of thin waist is achieved through the large deformation of the second area; on the other hand, the shape of the arm is maintained through the attenuation parameter of the first area; thus, the deformation effect of the entire image is improved.
  • the first area and the second area may be two adjacent areas.
  • the first area and the second area may be two separate areas.
  • a third area is provided between the first area and the second area; the second area is the area containing the second part that needs to be deformed; the first area contains the first part that needs to be deformed; the third area Since it does not include the area of the first part and the second part.
  • the position mapping formula when the grid points are deformed outside the first region is as follows:
  • src is the position of the grid point before deformation
  • dst is the position of the grid point after deformation
  • dst-src is the first displacement
  • src is the position of the grid point before deformation
  • dst is the position of the grid point after deformation
  • dst-src is the first displacement
  • s is the attenuation coefficient in the attenuation parameter.
  • the value range of s can be any value between 0 and 1. (dst-src)*(1-s) represents the second displacement smaller than the first displacement.
  • src is the position of the grid point before deformation
  • dst is the position of the grid point after deformation
  • dst-src is the first displacement
  • S is the attenuation value in the attenuation parameter.
  • the value range of S can be any positive integer.
  • src+(dst-src)-S represents a second displacement smaller than the first displacement.
  • an embodiment of the present disclosure also provides an image processing device, which includes:
  • the obtaining module 510 is configured to obtain the position information of the first key point of the first part contained in the target object in the first image;
  • the first determining module 520 is configured to determine a first area containing the first key point based on the location information of the first key point;
  • the second determining module 530 is configured to determine the displacement of the grid points in the first area according to the relative positions between the grid points of the deformed grid preset in the first area and the pixel points in the first area;
  • the control module 540 is configured to control the deformation of the pixel points in the first area according to the displacement of the grid points in the first area to obtain a second image after deformation.
  • the image processing apparatus provided in this embodiment is applied to various electronic devices that can be used for image deformation, for example, a terminal device or a server.
  • the acquisition module 510, the first determination module 520, the second determination module 530, and the control module 540 are all program modules. After the program modules are executed by the processor, the functions of any of the above modules can be realized.
  • the acquisition module 510, the first determination module 520, the second determination module 530, and the control module 540 are all software-hardware combined modules.
  • the software-hardware combined modules include but are not limited to programmable arrays; programmable arrays include But not limited to: Field Programmable Array and Complex Programmable Array.
  • the acquisition module 510, the first determination module 520, the second determination module 530, and the control module 540 are all pure hardware modules; the pure hardware modules include, but are not limited to, application specific integrated circuits.
  • the above-mentioned second determining module 530 is configured to determine the grid points of the grid points according to the relative positions between the grid points of the preset deformed grid in the first area and the pixel points in the first area.
  • the attenuation parameter of the displacement; and, according to the deformation command, the first displacement of the grid point is determined; the first displacement is attenuated according to the attenuation parameter to obtain a second displacement smaller than the first displacement.
  • control module 540 is configured to control the distance between adjacent pixels in the first area according to the second displacement amount to obtain the deformed second image.
  • the above-mentioned second determining module 530 is further configured to determine the pixel points located on the connecting line of the plurality of first key points to obtain the first set; according to the pixel points in the first set and the deformed grid intranet The relative positions of the grid points to obtain the second set, where the second set includes: the target grid points in the deformed grid with the closest distance to each pixel in the first set; according to each target grid point in the second set and the first set The relative positions of the pixel points in one set controlled by the target grid points determine the attenuation parameter of each target grid point in the second set.
  • the above-mentioned second determining module 530 is configured to traverse each target grid point in the second set outwards in a predetermined direction with each first key point as the center, to obtain each target in the second set
  • the grid points are sorted in a predetermined direction with respect to the spacing of each first key point; according to the spacing sort, the attenuation parameters of each target grid point in the second set are determined.
  • the above-mentioned second determining module 530 is configured to, in the case that any target grid point in the second set is located in a predetermined direction of the plurality of first key points, according to the corresponding information of the plurality of first key points Sort by spacing to determine the candidate value of the attenuation parameter; select the maximum value among the candidate values as the attenuation parameter of any target grid point.
  • the first part is the upper limb; the acquisition module 510 is configured to acquire the position information of the skeleton key points of the upper limb in the first image, and the skeleton key points include at least one of the following key points: shoulder key Points, elbow joint key points, wrist key points and hand key points.
  • the above-mentioned first determining module 520 is configured to determine the second area corresponding to the second part according to the position information of the second key point of the second part contained in the target in the first image;
  • control module 540 is further configured to control the deformation of the second area in the first image according to the first displacement of the grid points of the preset deformed grid in the second area to obtain the deformed second image.
  • Grid deformation is a deformation tool used for image deformation.
  • the deformed grid before deformation is a regular grid, usually including straight warp and latitude lines to form a rectangular grid.
  • the first area and the second area are determined on the image to be deformed.
  • the first area is the area where the first part where the deformation amplitude needs to be suppressed is located
  • the second area is the area where the second part where the deformation amplitude does not need to be suppressed relative to the first area.
  • the arm may be beside the waist.
  • the technical solution of the embodiment of the present disclosure is adopted, and the arm may be the aforementioned first part;
  • the waist may be the aforementioned second part.
  • the attenuation processing is performed on the basis of the first displacement of the grid points that fall in the first area (for example, the first displacement may be the original displacement directly obtained based on the deformation command) to obtain the ratio
  • the second displacement is smaller than the first displacement.
  • the image processing method provided in this example may include:
  • the first step is to determine the first key point; for example, determining the first key point may include: determining 4 key points of the arm as the first key point. Assume that these four first key points are referred to as the four key points of ABCD; for example, for the upper limbs, the four key points of ABCD can be: shoulder key points, elbow joint key points, wrist key points, and hand key points;
  • the second step is to connect the four key points ABCD to get the line formed by the first key point;
  • the third step is to determine the first set and the second set; the pixels included in the first set are located on the above-mentioned line.
  • the grid points contained in the second set are: the grid points in the deformed grid with the smallest distance from the pixel points in the first set.
  • the fourth step is to determine the attenuation parameter of each grid point in the second set according to the relative position of each grid point in the second set and the pixel point controlled by the corresponding grid point in the first set;
  • the first displacement can be attenuated according to the attenuation parameter of each grid point to obtain a second displacement smaller than the first displacement, and the deformation processing is performed based on the second displacement;
  • the deformation processing can be performed directly based on the first displacement amount.
  • the deformation amplitude of the first area is smaller than the deformation amplitude of the second area, so as to achieve fine control of the deformation of pixels in different areas in the same image, thereby improving Image distortion effect.
  • an embodiment of the present disclosure also provides an image processing device, including:
  • Memory used to store information
  • the processor is respectively connected to the display and the memory, and is used to execute the computer executable instructions stored on the memory to implement the image processing method provided by one or more technical solutions, such as those shown in FIG. 1 and/or FIG. 3 Image processing method.
  • the memory can be various types of memory, such as random access memory, read-only memory, flash memory, and so on.
  • the memory can be used for information storage, for example, to store computer-executable instructions.
  • the computer-executable instructions may be various program instructions, for example, target program instructions and/or source program instructions.
  • the processor may be various types of processors, for example, a central processing unit, a microprocessor, a digital signal processor, a programmable array, a digital signal processor, an application specific integrated circuit, or an image processor.
  • the processor can be connected to the memory via a bus.
  • the bus may be an integrated circuit bus or the like.
  • the terminal device may further include: a communication interface, and the communication interface may include: a network interface; the network interface includes, for example, a local area network interface, a transceiver antenna, and the like.
  • the communication interface is also connected to the processor and can be used to send and receive information.
  • the terminal device further includes a human-computer interaction interface.
  • the human-computer interaction interface may include various input and output devices, such as a keyboard, a touch screen, and the like.
  • the image processing device further includes a display, which can display various prompts, collected facial images, and/or various interfaces.
  • the embodiments of the present disclosure also provide a computer storage medium, the computer storage medium stores computer executable code; after the computer executable code is executed, the image processing method provided by one or more technical solutions can be implemented, such as FIG. 1 and / Or the image processing method shown in Figure 3.
  • the disclosed device and method may be implemented in other ways.
  • the device embodiments described above are only illustrative.
  • the division of the units is only a logical function division, and there may be other divisions in actual implementation, such as: multiple units or components can be combined, or It can be integrated into another system, or some features can be ignored or not implemented.
  • the coupling, or direct coupling, or communication connection between the components shown or discussed may be indirect coupling or communication connection through some interfaces, devices or units, and may be in electrical, mechanical or other forms. of.
  • the units described above as separate components may or may not be physically separate, and the components displayed as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units; Some or all of the units may be selected according to actual needs to achieve the objectives of the solutions of the embodiments.
  • the functional units in the embodiments of the present disclosure can be all integrated into one processing module, or each unit can be individually used as a unit, or two or more units can be integrated into one unit;
  • the unit can be implemented in the form of hardware, or in the form of hardware plus software functional units.
  • a person of ordinary skill in the art can understand that all or part of the steps in the above method embodiments can be implemented by a program instructing relevant hardware.
  • the foregoing program can be stored in a computer readable storage medium. When the program is executed, it is executed. Including the steps of the foregoing method embodiment; and the foregoing storage medium includes: removable storage devices, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), magnetic disks or optical disks, etc.
  • ROM read-only memory
  • RAM Random Access Memory
  • magnetic disks or optical disks etc.

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Multimedia (AREA)
  • Human Computer Interaction (AREA)
  • Image Processing (AREA)
  • Processing Or Creating Images (AREA)
  • Image Analysis (AREA)

Abstract

本公开实施例公开一种图像处理方法及装置、图像处理设备及存储介质。所述图像处理方法，包括：获取第一图像中目标对象包含的第一部位的第一关键点的位置信息；基于所述第一关键点的位置信息，确定包含所述第一关键点的第一区域；根据所述第一区域内预设的变形网格的网格点与所述第一区域内的像素点之间的相对位置，确定所述第一区域内网格点的位移量；根据所述第一区域内网格点的位移量，控制所述第一区域内像素点的变形，得到变形后的第二图像。

Description

图像处理方法及装置、图像处理设备及存储介质
相关申请的交叉引用
本公开基于申请号为201911360894.4、申请日为2019年12月25日的中国专利申请提出，并要求该中国专利申请的优先权，该中国专利申请的全部内容在此以引入方式并入本公开。
技术领域
本公开涉及图像技术领域，尤其涉及一种图像处理方法及装置、图像处理设备及存储介质。
背景技术
在图像技术领域，存在着对用户拍完一个照片，然后需要对该照片的部分进行变形处理。目前，通常会对整张图像进行变形，或者对图像中的部分区域进行变形、而对其他区域不进行变形处理。针对图像中的部分区域进行变形处理的情况，经过变形处理的区域和未经变形处理的区域之间的过渡很不自然，进而使得图像效果很差，并不能使得用户满意。
发明内容
本公开实施例期望提供一种图像处理方法及装置、图像处理设备及存储介质。
本公开实施例的技术方案是这样实现的：
本公开实施例第一方面提供一种图像处理方法，包括：获取第一图像中目标对象包含的第一部位的第一关键点的位置信息；基于所述第一关键点的位置信息，确定包含所述第一关键点的第一区域；根据所述第一区域内预设的变形网格的网格点与所述第一区域内的像素点之间的相对位置，确定所述第一区域内网格点的位移量；根据所述第一区域内网格点的位移量，控制所述第一区域内像素点的变形，得到变形后的第二图像。
在本公开的一些可选实施例中，所述根据所述第一区域内预设的变形网格的网格点与所述第一区域内的像素点之间的相对位置，确定所述第一区域内网格点的位移量，包 括：根据所述第一区域内的预设的变形网格的网格点与所述第一区域内的像素点之间的相对位置，确定所述网格点的位移量的衰减参数；以及，根据变形指令，确定所述网格点的第一位移量；根据所述衰减参数对所述第一位移量进行衰减处理，得到比所述第一位移量小的第二位移量。
在本公开的一些可选实施例中，所述根据所述第一区域内网格点的位移量，控制所述第一区域内像素点的变形，得到变形后的第二图像，包括：根据所述第二位移量，控制所述第一区域内的相邻像素点之间的间距，得到变形后的第二图像。
在本公开的一些可选实施例中，所述根据所述第一区域内预设的变形网格的网格点与所述第一区域内像素之间的相对位置，确定所述网格点的位移量的衰减参数，包括：确定位于多个所述第一关键点的连线上的像素点，得到第一集合；根据所述第一集合内像素点与所述变形网格内网格点的相对位置，得到第二集合，其中，所述第二集合包括：所述变形网格中与所述第一集合内各像素点距离最近的目标网格点；根据所述第二集合中各所述目标网格点与所述第一集合中由所述目标网格点所控制的像素点之间的相对位置，确定所述第二集合中各所述目标网格点的所述衰减参数。
在本公开的一些可选实施例中，所述根据所述第二集合中各所述目标网格点与所述第一集合中由所述目标网格点所控制的像素点之间的相对位置，确定所述第二集合中各所述网格点的所述衰减参数，包括：分别以各所述第一关键点为中心，在预定方向上向外进行遍历所述第二集合中各所述目标网格点，得到所述第二集合中各所述目标网格点相对于每个所述第一关键点在预定方向上的间距排序；根据所述间距排序，确定所述第二集合中各所述目标网格点的所述衰减参数。
在本公开的一些可选实施例中，所述根据所述间距排序，确定所述第二集合中各所述目标网格点的所述衰减参数，包括：在所述第二集合中任意一个所述目标网格点位于多个所述第一关键点的预定方向的情况下，根据多个所述第一关键点所对应的所述间距排序，确定所述衰减参数的备选值；选择所述备选值中的最大值，作为任意一个所述目标网格点的所述衰减参数。
在本公开的一些可选实施例中，所述第一部位为上肢；所述获取第一图像中目标所包含第一部位的关键点的位置信息，包括：获取所述第一图像中上肢的骨架关键点的位置信息，所述骨架关键点包括以下关键点中的至少一种：肩部关键点、肘关节关键点、手腕关键点和手部关键点。
在本公开的一些可选实施例中，所述方法还包括：根据所述第一图像中目标所包含 的第二部位的第二关键点的位置信息，确定第二部位对应的第二区域；根据所述第二区域内的预设的变形网格的网格点的第一位移量，控制所述第一图像内第二区域的变形，得到变形后的第二图像。
本公开实施例公开的一种图像处理装置，包括：获取模块，配置为获取第一图像中目标对象包含的第一部位的第一关键点的位置信息；第一确定模块，配置为基于所述第一关键点的位置信息，确定包含所述第一关键点的第一区域；第二确定模块，配置为根据所述第一区域内预设的变形网格的网格点与所述第一区域内的像素点之间的相对位置，确定所述第一区域内网格点的位移量；控制模块，配置为根据所述第一区域内网格点的位移量，控制所述第一区域内像素点的变形，得到变形后的第二图像。
在本公开的一些可选实施例中，所述第二确定模块，配置为根据所述第一区域内的预设的变形网格的网格点与所述第一区域内的像素点之间的相对位置，确定所述网格点的位移量的衰减参数；以及，根据变形指令，确定所述网格点的第一位移量；根据所述衰减参数对所述第一位移量进行衰减处理，得到比所述第一位移量小的第二位移量。
在本公开的一些可选实施例中，所述控制模块，配置为根据所述第二位移量，控制所述第一区域内的相邻像素点之间的间距，得到变形后的第二图像。
在本公开的一些可选实施例中，所述第二确定模块，配置为确定位于多个所述第一关键点的连线上的像素点，得到第一集合；根据所述第一集合内像素点与所述变形网格内网格点的相对位置，得到第二集合，其中，所述第二集合包括：所述变形网格中与所述第一集合内各像素点距离最近的目标网格点；根据所述第二集合中各所述目标网格点与所述第一集合中由所述目标网格点所控制的像素点之间的相对位置，确定所述第二集合中各所述目标网格点的所述衰减参数。
在本公开的一些可选实施例中，所述第二确定模块，配置为分别以各所述第一关键点为中心，在预定方向上向外进行遍历所述第二集合中各所述目标网格点，得到所述第二集合中各所述目标网格点相对于每个所述第一关键点在预定方向上的间距排序；根据所述间距排序，确定所述第二集合中各所述目标网格点的所述衰减参数。
在本公开的一些可选实施例中，所述第二确定模块，配置为在所述第二集合中任意一个所述目标网格点位于多个所述第一关键点的预定方向的情况下，根据多个所述第一关键点所对应的所述间距排序，确定所述衰减参数的备选值；选择所述备选值中的最大值，作为任意一个所述目标网格点的所述衰减参数。
在本公开的一些可选实施例中，所述第一部位为上肢；所述获取模块，配置为获取 所述第一图像中上肢的骨架关键点的位置信息，所述骨架关键点包括以下关键点中的至少一种：肩部关键点、肘关节关键点、手腕关键点和手部关键点。
在本公开的一些可选实施例中，所述第一确定模块，还配置为根据所述第一图像中目标所包含的第二部位的第二关键点的位置信息，确定第二部位对应的第二区域；
所述控制模块，还配置为根据所述第二区域内的预设的变形网格的网格点的第一位移量，控制所述第一图像内第二区域的变形，得到变形后的第二图像。
本公开实施例第三方面提供一种图像处理设备，包括：存储器；处理器，与所述存储器连接，用于通过执行存储在所述存储器上的计算机可执行指令实现前述任意技术方案提供的图像处理方法。
本公开实施例第四方面提供一种计算机存储介质，所述计算机存储介质存储有计算机可执行指令；所述计算机可执行指令被处理器执行后，能够实现前述任意技术方案提供的图像处理方法。
本公开实施例提供的技术方案，在利用变形网格对整个第一图像进行变形之前，可以先确定出第一部位的第一关键点，然后基于第一关键点得到需要保护的第一区域内的像素点，在进行变形时，第一区域内网格点的位移量是根据该第一区域内的网格点与像素点之间的相对位置来确定的，而不再是仅仅基于单一的变形指令来确定。如此，引入网格点与像素点之间的相对位置来确定网格点的位移量，从而可以实现对第一区域内网格点的位移量的精确控制，进而可以实现对同一张图像内不同区域内像素的变形的精细控制，从而有效提升图像变形效果。
附图说明
图1为本公开实施例提供的一种图像处理方法的流程示意图；
图2A为本公开实施例提供的一种铺设在第一图像上的标准变形网格的使用图；
图2B为本公开实施例提供的一种第一区域和第二区域的示意图；
图3为本公开实施例提供的一种确定衰减参数的流程示意图；
图4为本公开实施例提供的一种关键点的连线及第二集合的网格点的示意图；
图5为本公开实施例提供的一种图像处理装置的结构示意图；
图6为本公开实施例提供的一种图像处理设备的结构示意图。
具体实施方式
以下结合说明书附图及具体实施例对本公开实施例的技术方案做进一步的详细阐述。
在本公开实施例的术语是仅仅出于描述特定实施例的目的，而非旨在限制本公开实施例。在本公开实施例和所附权利要求书中所运行的单数形式的“一种”、“所述”和“该”也旨在包括多数形式，除非上下文清楚地表示其他含义。还应当理解，本公开实施例中的术语“和/或”是指并包含一个或多个相关联的列出项目的任何或所有可能组合。
应当理解，尽管在本公开实施例可能采用术语第一、第二、第三等来描述各种信息，但这些信息不应限于这些术语。这些术语仅用来将同一类型的信息彼此区分开。例如，在不脱离本公开实施例范围的情况下，第一信息也可以被称为第二信息，类似地，第二信息也可以被称为第一信息。取决于语境，如在此所运行的词语“如果”可以被解释成为“在……时”或“当……时”或“响应于确定”。
如图1所示，本公开实施例提供一种图像处理方法，包括：
步骤S110：获取第一图像中目标对象包含的第一部位的第一关键点的位置信息；
步骤S120：基于第一关键点的位置信息，确定包含第一关键点的第一区域；
步骤S130：根据第一区域内预设的变形网格的网格点与第一区域内的像素点之间的相对位置，确定第一区域内网格点的位移量；
步骤S140：根据第一区域内网格点的位移量，控制第一区域内像素点的变形，得到变形后的第二图像。
本实施例提供的图像处理方法，可以应用于具有图像处理功能的电子设备中。示例性的，该图像设备可包括各种终端设备，该终端设备包括：手机或可穿戴式设备等。该终端设备还可包括：车载终端设备，或专用于图像采集且固定于某一处的固定终端设备。在另一些实施例中，图像设备还可包括：服务器，例如，本地服务器或者位于云平台中提供图像处理服务的云服务器等。
在一些实施例中，目标对象例如为人体、动物、或者由虚拟三维模型渲染而成的虚拟对象等，本公开对目标对象的具体形式并不限定。目标对象的第一部位可以是肢体部位，例如，对于目标对象为人体的情况，目标对象的第一部位，可以是手臂、腿部或者腹部等，本公开实施例对此并不限定。
本实施例中，在进行图像的变形处理之前，第一图像可以分为多个区域，第一区域 可包括多个区域中的一个或多个区域。
在一些实施例中，第一区域可以是包含需要抑制变形的第一部位的区域，或者，第一区域可以是包含需要增强变形的第一部位的区域。示例性的，需要增强变形的第一部位的区域可以为需变形的程度比较大的区域；需要抑制变形的第一部位的区域可以为需变形的程度比较小的区域。
本实施例中，在获取到第一图像之后，对第一图像进行变形处理之前，可确定变形网格。例如，在第一图像上铺设变形网格。示例性的，变形网格可包括：横线和纵线交叉形成的网格点。变形网格所包含的横线可以称之为纬线，纵线可以称之为经线；变形网格内的线可统称为经纬线。在对第一图像进行变形处理前，经纬线可以是标准的横纵方向分别铺设的直线。
若对由经纬线构成的变形网格进行统一变形，则图像中的各区域的变形幅度是相同的，这样可能会使得无需变形、需变形的程度较小或需变形的程度较大的区域也会被无差别的按照统一的变形幅度进行变形，并且这种按照统一的变形幅度的变形方式，可能会使得生成的第二图像出现不协调等现象，从而导致第一图像的变形效果不佳。
图2A为一张铺设有预设的变形网格的示意图。图2B是以图2A中的人像为例，确定右侧上肢区域作为第一区域。
在本实施例中，首先确定出第一图像中目标所包含第一部位的第一关键点。示例性的，该第一关键点可以为第一部位的骨架关键点或轮廓关键点，骨架关键点为人体骨骼或者动物骨骼所在位置的关键点。轮廓关键点可为人体或动物体表所呈现轮廓的关键点。可以理解，第一关键点可以是位于第一部位上的点，且可以用于定位第一部位的点，骨架关键点的分布位置决定了第一部位在第一图像中的所在位置。故在本实施例中，可以基于一个或多个骨架关键点所在位置，确定第一区域。
本实施例中，步骤S120中，可基于第一关键点的位置信息，至少确定出第一区域的边界；基于第一区域的边界可确定第一区域。在步骤S130中，确定出第一区域后，可得到第一区域内的像素点。
本实施例中，利用预设的变形网格进行图像变形的变形方法可以称为网格变形。在一些实施例中，在利用预设的变形网格进行图像变形时，第一区域内的像素点是可能要移动位置的像素点，移动像素点的位置之后，像素点之间的间距大小会发生变化。
在一些实施例中，S140可包括：根据第一区域内网格点的位移量，调整第一区域内像素点之间的疏密程度，从而实现控制第一区域内像素点的变形，得到变形后的第二 图像。
如此，确定出第一图像中的第一部位的关键点，再基于第一关键点得到第一区域内的像素点；在进行图像变形时，第一区域内网格点的位移量是基于该区域内的网格点与像素点之间的相对位置确定的，而不再是仅仅基于单一的变形指令确定。如此，可以通过对第一区域内网格点的位移量的精确控制，实现图像的不同区域内像素变形(即像素点之间的间距大小)的精细控制，从而提升图像变形效果。
在本公开的一些可选实施例中，S130可包括：根据第一区域内的预设的变形网格的网格点与第一区域内的像素点之间的相对位置，确定网格点的位移量的衰减参数；以及，根据变形指令，确定网格点的第一位移量；根据衰减参数对第一位移量进行衰减处理，得到比第一位移量小的第二位移量。
在本实施例中，可以根据网格点与第一区域内的像素点之间的相对位置得到第一区域内的特定像素与各网格点之间的距离，进而根据该距离确定网格点的位移量的衰减参数的大小。例如，该特定像素点可为第一关键点所在位置的像素点，或者，该特定像素点可为第一关键点所在位置附近的像素点。此处仅是一种基于上述相对位置，确定第一区域内的网格点的位移量的衰减参数的一种举例，具体不限于此。
本实施例中，变形指令可以是根据人机交互界面接收的用户输入而生成的指令，或者，基于一键美容或美体等图像预设处理功能生成的变形指令。例如，具有针对图像中的人像的自动瘦腰功能，则图像处理设备内会根据自动瘦腰功能生成对应的变形指令。该变形指令可以携带有变形参数，示例性的，该变形参数可以包括上述第一位移量。
在确定第一位移量之后，利用衰减参数及第一位移量作为已知量，进行针对各第一位移量的衰减处理，从而得到比第一位移量小的第二位移量。
在本公开的一些可选实施例中，衰减参数为用于减小第一区域内网格点位移量的参数。第一区域内像素点的变形幅度与第一区域内网格点的位移量是正相关的，即，网格点的位移量越大，则第一区域内像素点的变形幅度越大；相应的，网格点的位移量越小，则第一区域内像素点的变形幅度越小。
在一些可选实施例中，衰减参数包括但不限于以下至少之一：
衰减系数，衰减系数也可以称为衰减比值；例如，根据变形指令计算得到第一区域内各网格点原始的第一位移量，将该第一位移量与上该衰减系数相乘，得到第一区域内各网格点最终的第二位移量。
衰减值，若该衰减值为正值，则可以将原始的第一位移量减去该衰减值，得到比原 始的第一位移量更小的第二位移量。
在本公开的一些可选实施例中，步骤S140可包括：根据第二位移量，控制第一区域内的相邻像素点之间的间距，得到变形后的第二图像。
示例性的，根据第二位移量增大第一区域内某些相邻像素点之间的间距，和/或，根据第二位移量缩小第一区域内某些相邻像素点之间的间距，从而变形后的第一区域由相邻像素点的等间距的特点转变为相邻像素点之间的不等间距的特点。例如，网格点A的第二位移量大于网格点B的第二位移量，则网格点A所控制的像素点之间的间距改变量，大于网格点B所控制的像素点之间的间距改变量。
例如，若腰部所在区域所包含相邻像素点之间的间距缩小了，在腰部所对应像素点总个数不变的情况下，达到了瘦腰的效果。若胸部所在区域所包含相邻像素点之间的间距增大了，则在胸部所对应像素点总个数不变的情况下，达到丰胸的效果。
如此，第一区域内相邻像素点之间的间距变化越大，则对应部位或区域的变形幅度越大。反之，第一区域内相邻像素点之间的间距变化越小，则对应部位或区域的变形幅度也越小。
本实施例中，通过对第一区域内网格点的位移量做衰减处理，将基于变形指令得到的第一位移量缩小为第二位移量，如此，能够抑制(即弱化)第一区域的变形幅度，从而使得第一区域和其他区域的变形幅度不同，满足不同区域不同变形幅度的变形需求，从而提升了第一图像变形为第二图像的变形效果。
在本公开的一些可选实施例中，如图3所示，上述根据第一区域内的预设的变形网格的网格点与第一区域内的像素点之间的相对位置，确定网格点的位移量的衰减参数，可包括如下步骤：
步骤301：确定位于多个第一关键点的连线上的像素点，得到第一集合；
步骤302：根据第一集合内像素点与变形网格内网格点的相对位置，得到第二集合，其中，第二集合包括：变形网格中与第一集合内各像素点距离最近的目标网格点；
步骤303：根据第二集合中各目标网格点与第一集合中由目标网格点所控制的像素点之间的相对位置，确定第二集合中各目标网格点的衰减参数。
本实施例中，确定出第一关键点之后，直接连线相邻的第一关键点得到一条或多条连线，位于这些连线上的像素点构成第一集合。参考图4所示，第一关键点连线上的像素点构成第一集合。得到第一集合后，根据变形网格中各个网格点与第一集合中像素点的相对位置，找到分别与第一集合中每个像素点距离最近的网格点作为目标网格点，构 成第二集合。
其中，第一集合是第一关键点的连线上的像素点形成的；第二集合包含与上述连线相邻的目标网格点。故第一集合是像素点的集合，第二集合是网格点的集合，具体是目标网格点的集合。
由于在变形过程中，第一图像的变形是基于变形网格的网格点进行的。第二集合中目标网格点所围成的区域即为上述第一区域。
本实施例中，第一集合中由目标网格点所控制的像素点，可以是与目标网格点距离最近的像素点。可以理解，第二集合中每一个目标网格点的衰减参数，可以根据各个目标网格点与第一关键点所形成连线上像素点之间的相对位置确定的。
示例性的，一方面，第二集合中的目标网格点，有的靠近第一关键点，有的远离第一关键点。第二集合中靠近第一关键点的目标网格点，相对于远离第一关键点目标网格点，具有更大的衰减参数。另一方面，第一集合内的像素点可为第一区域内的部分像素点。在第一区域中第一集合以外的剩余像素点，有的像素点靠近第一集合中的像素点，有的像素点远离第一集合中的像素点。远离第一集合中像素点的像素点，一般由远离第一关键点的目标网格点控制，因此远离第一关键点的像素点衰减的变形幅度，小于靠近第一关键点的像素点衰减的变形幅度。
在本公开的一些可选实施例中，步骤303可包括：分别以各第一关键点为中心，在预定方向上向外进行遍历第二集合中各目标网格点，得到第二集合中各目标网格点相对于每个第一关键点在预定方向上的间距排序；根据间距排序，确定第二集合中各目标网格点的衰减参数。
在一些实施例中，可以根据第二集合中各目标网格点，与第一集合中各像素点的最小距离，确定第二集合内各目标网格点的衰减参数。上述最小距离即为上述间距排序中所使用的间距的一种可选举例。例如，假设第一集合中有M个像素点，第二集合有N个目标网格点。M，N均为正整数。这N个目标网格点与M个像素点具有M个距离，确定N个目标网格点中每一个目标网格点所对应的M个距离中最小距离。将N个目标网格点所对应的最小距离进行排序，得到上述间距排序。上述最小距离越小，对应的间距排序越靠前，对应的目标网格点的衰减参数越大；相应的，上述最小距离越大，对应的间距排序越靠后，对应的目标网格点的衰减参数越小。如此，网格点与第一关键点之间的相对位置，可以用上述间距或者间距排序来表征。
但是这种方式在第一部位的姿势比较特殊的时候，可能会出现离第一部位的中心位 置最近的网格点衰减参数反而越小的情况，从而导致出现图像效果不如预期的现象。
在本实施例中，以各第一关键点为遍历的起始位置，通过遍历第二集合中每一个目标网格点，得到第二集合中每一个目标网格点相对于该第一关键点在预定方向上的间距排序。示例性的，该预定方向可以为：第一部位需要抑制变形的方向、或者禁止变形的方向。在一些实施例中，该预定方向可以为：变形网格的横线方向和纵线方向中，与对应第一关键点的连线夹角较大的方向。在另一些实施例中，该预定方向可以为：变形网格的横线方向和纵线方向中，与第一关键点所形成连线的整体延伸方向夹角较大的方向。在又一些实施例中，该预定方向可以为：变形网格的横线方向和纵线方向中与第一部位的延伸方向夹角较大的方向。在通常情况下，第二集合中一个网格点可能仅位于一个像素点的预定方向上。
本实施例中，上述间距排序可与对应的衰减参数具有一定的相关性。例如，若衰减参数直接用于目标网格点的第一位移量的衰减，则间距排序越靠前，则衰减参数的值越小。
在一些可选实施例中，上述根据间距排序，确定第二集合中目标网格点的衰减参数，包括：在第二集合中任意一个目标网格点位于多个第一关键点的预定方向的情况下，根据各个第一关键点所对应的间距排序，确定衰减参数的备选值；选择备选值中的最大值，作为任意一个目标网格点的衰减参数。
本实施方式中，如果第二集合中一个目标网格点位于多个第一关键点的预定个方向时，会确定出多个间距排序。此时，一个间距排序对应一个衰减参数。这样可得到一个目标网格点对应于不同第一关键点的间距排序，进而可得到多个衰减参数的备选参值。最终选择多个衰减参数的备选值中的最大值，作为该目标网格点的衰减参数，如此，确保了离第一关键点所形成连线越近的目标网格点的衰减参数越大的情况。
在本公开的一些可选实施例中，第一部位为上肢。该上肢可包括：上臂、前臂和/或手部的其中至少之一。
在本公开的一些可选实施例中，S110可包括：获取第一图像中上肢的骨架关键点的位置信息，骨架关键点包括以下关键点中的至少一种：肩部关键点、肘关节关键点、手腕关键点和手部关键点。则上述第一关键点的连线可为：从肩部关键点、肘关节关键点、手腕关键点到手部关键点中的至少一种关键点的依次连线。则前述第一集合包含的像素点可以包括以下至少之一：肩部关键点到肘关节关键点之间连线上的像素点、肘关节关键点到手腕关键点之间连线上的像素点、手腕关键点到手部关键点之间连线上的像素 点。
在本公开的一些可选实施例中，方法还包括：根据第一图像中目标所包含的第二部位的第二关键点的位置信息，确定第二部位对应的第二区域；根据第二区域内的预设的变形网格的网格点的第一位移量，控制第一图像内第二区域的变形，得到变形后的第二图像。其中，第一部位不同于第一部位。参照图2B所示，还可以确定除右侧上肢以外的腰部区域作为第二区域。
在本公开实施例中，第一区域内和第二区域内的网格点的第一位移量，均可以为根据变形指令确定的初始位移量。
变形幅度可由网格点的位移量控制。故在本实施例中，第一区域可以是需要抑制变形的区域，而第二区域可以是需要进行变形的区域。在利用预设的变形网格进行变形过程中，可以基于衰减参数，在基于相同的变形指令的情况下，使得第一区域的变形幅度小于第二区域的变形幅度。其中，变形幅度对应的变形方向包括但不限于以下至少之一：对应区域的变形部位的增大、缩小、旋转、镜像、线条形状的改变等。
例如，在针对图像中的人体进行变形处理时，若变形处理的第一部位为腰部，在利用预设的变形网格进行收腰变形处理过程中，向人像的中心压缩腰部时，位于腰部附近的手臂可能会发生拉伸的变形。
为了减少因为腰部变形对手臂的负面影响，在本实施例中，可以将手臂所在的图像区域设置为第一区域，而将腰部所在的图像区域为第二区域。如此，在本实施例中通过衰减参数使得第一区域和第二区域利用同一个变形网格进行变形时，使得第一区域的变形幅度小，而第二区域的变形幅度大；如此，一方面通过第二区域的大幅度变形实现瘦腰的效果；另一方面，通过第一区域的衰减参数维持手臂的形状；从而提升整个图像的变形效果。
在一些实施例中，第一区域和第二区域可以为相邻的两个区域。
在另一些实施例中，第一区域和第二区域可以为分离的两个区域。例如，在第一区域和第二区域之间设置有第三区域；第二区域为包含需要变形的第二部位的区域；第一区域包含有需要抑制变形的第一部位的区域；第三区域为既然不包含第一部位和第二部位的区域。
在一些实施例中，第一区域外(例如，第二区域)内网格点变形时的位置映射公式如下：
src+(dst-src)             (1)
其中，src为网格点变形前的位置；dst为网格点变形后的位置；dst-src为第一位移量。
第一区域内的网格点变形时采用的第二位移量的计算公式(2)或公式(3)：
src+(dst-src)*(1-s)         (2)
其中，src为网格点变形前的位置；dst为网格点变形后的位置；dst-src为第一位移量，s为衰减参数中的衰减系数。示例性的，s的取值范围可以为0到1之间的任意数值。(dst-src)*(1-s)表示比第一位移量小的第二位移量。
src+(dst-src)-S             (3)
其中，src为网格点变形前的位置；dst为网格点变形后的位置；dst-src为第一位移量，S为衰减参数中的衰减值。示例性的，S的取值范围可以为任意正整数。src+(dst-src)-S表示比第一位移量小的第二位移量。
如图5所示，本公开实施例还提供一种图像处理装置，装置包括：
获取模块510，配置为获取第一图像中目标对象包含的第一部位的第一关键点的位置信息；
第一确定模块520，配置为基于第一关键点的位置信息，确定包含第一关键点的第一区域；
第二确定模块530，配置为根据第一区域内预设的变形网格的网格点与第一区域内的像素点之间的相对位置，确定第一区域内网格点的位移量；
控制模块540，配置为根据第一区域内网格点的位移量，控制第一区域内像素点的变形，得到变形后的第二图像。
本实施例提供的图像处理装置应用于能够用于图像变形的各种电子设备中，例如，终端设备或者服务器中。
在一些实施例中，上述获取模块510、第一确定模块520、第二确定模块530及控制模块540均为程序模块，程序模块被处理器执行后，能够实现上述任意模块的功能。
在另一些实施例中，上述获取模块510、第一确定模块520、第二确定模块530及控制模块540均为软硬结合模块，软硬结合模块包括但不限于可编程阵列；可编程阵列包括但不限于：现场可编程阵列和复杂可编程阵列。
在又一些实施例中，上述获取模块510、第一确定模块520、第二确定模块530及控制模块540均为纯硬件模块；纯硬件模块包括但不限于专用集成电路。
在一些实施例中，上述第二确定模块530，配置为根据第一区域内的预设的变形网 格的网格点与第一区域内的像素点之间的相对位置，确定网格点的位移量的衰减参数；以及，根据变形指令，确定网格点的第一位移量；根据衰减参数对第一位移量进行衰减处理，得到比第一位移量小的第二位移量。
在一些实施例中，上述控制模块540，配置为根据第二位移量，控制第一区域内的相邻像素点之间的间距，得到变形后的第二图像。
在一些实施例中，上述第二确定模块530，还配置为确定位于多个第一关键点的连线上的像素点，得到第一集合；根据第一集合内像素点与变形网格内网格点的相对位置，得到第二集合，其中，第二集合包括：变形网格中与第一集合内各像素点距离最近的目标网格点；根据第二集合中各目标网格点与第一集合中由目标网格点所控制的像素点之间的相对位置，确定第二集合中各目标网格点的衰减参数。
在一些实施例中，上述第二确定模块530，配置为分别以各第一关键点为中心，在预定方向上向外进行遍历第二集合中各目标网格点，得到第二集合中各目标网格点相对于每个第一关键点在预定方向上的间距排序；根据间距排序，确定第二集合中各目标网格点的衰减参数。
在一些实施例中，上述第二确定模块530，配置为在第二集合中任意一个目标网格点位于多个第一关键点的预定方向的情况下，根据多个第一关键点所对应的间距排序，确定衰减参数的备选值；选择备选值中的最大值，作为任意一个目标网格点的衰减参数。
在一些实施例中，上述第一部位为上肢；上述获取模块510，配置为获取第一图像中上肢的骨架关键点的位置信息，骨架关键点包括以下关键点中的至少一种：肩部关键点、肘关节关键点、手腕关键点和手部关键点。
在一些实施例中，上述第一确定模块520，配置为根据第一图像中目标所包含的第二部位的第二关键点的位置信息，确定第二部位对应的第二区域；
上述控制模块540，还配置为根据第二区域内的预设的变形网格的网格点的第一位移量，控制第一图像内第二区域的变形，得到变形后的第二图像。
以下为以上实施例的具体示例性说明：
网格变形是用于图像变形的一种变形工具，变形前的变形网格是规整的网格，通常是包括笔直的经线和纬线构成矩形网格。
在待变形的图像上确定出第一区域和第二区域。第一区域为需要抑制变形幅度的第一部位所在区域，而第二区域相对于第一区域是无需抑制变形幅度的第二部位所在区域。例如，利用变形网格对人体的腰部进行收腰变形处理时，手臂可能处于 腰部旁边，为了减少腰部变形导致对手臂的影响，采用本公开实施例的技术方案，手臂可为前述第一部位；腰部可为前述的第二部位。
在本示例中，针对落在第一区域里的网格点的第一位移量(例如，该第一位移量可为基于变形指令直接得到的原始位移量)的基础上进行衰减处理，得到比第一位移量小的第二位移量。
如此，通过减少第一区域里的网格点的位移量，就能够减小第一区域内的像素点的变形幅度。
以手臂为前述第一部位、以包含手臂的图像区域为第一区域为例进行举例说明，则本示例提供的图像处理方法可包括：
第一步，确定第一关键点；例如，确定第一关键点可包括：确定手臂的4个关键点作为第一关键点。假设这4个第一关键点简称ABCD四个关键点；例如，针对上肢而言，ABCD四个关键点可为：肩部关键点、肘关节关键点、手腕关键点及手部关键点；
第二步，连接ABCD四个关键点，得到第一关键点所形成的连线；
第三步，确定第一集合和第二集合；第一集合内包括的像素点位于上述连线上。第二集合包含的网格点为：变形网格中与第一集合内像素点距离最小的网格点。
第四步，根据第二集合中各网格点与第一集合中对应网格点所控制像素点的相对位置，确定第二集合中各网格点的衰减参数；
第五步，针对第一区域，可以根据各网格点的衰减参数对第一位移量进行衰减处理，得到比第一位移量小的第二位移量，并基于第二位移量进行变形处理；针对第二区域，可以直接基于第一位移量进行变形处理。
从而可以通过对第一区域内网格点的位移量的精确控制，使得第一区域的变形幅度小于第二区域的变形幅度，实现同一张图像内不同区域内像素的变形的精细控制，从而提升图像变形效果。
如图6所示，本公开实施例还提供了一种图像处理设备，包括：
存储器，用于存储信息；
处理器，分别与显示器及存储器连接，用于通过执行存储在存储器上的计算机可执行指令，能够实现前述一个或多个技术方案提供的图像处理方法，例如图1和/或图3所示的图像处理方法。
该存储器可为各种类型的存储器，可为随机存储器、只读存储器、闪存等。存 储器可用于信息存储，例如，存储计算机可执行指令等。计算机可执行指令可为各种程序指令，例如，目标程序指令和/或源程序指令等。
处理器可为各种类型的处理器，例如，中央处理器、微处理器、数字信号处理器、可编程阵列、数字信号处理器、专用集成电路或图像处理器等。
处理器可以通过总线与存储器连接。总线可为集成电路总线等。
在一些实施例中，终端设备还可包括：通信接口，该通信接口可包括：网络接口；网络接口例如包括局域网接口、收发天线等。通信接口同样与处理器连接，能够用于信息收发。
在一些实施例中，终端设备还包括人机交互接口，例如，人机交互接口可包括各种输入输出设备，例如，键盘、触摸屏等。
在一些实施例中，图像处理设备还包括：显示器，该显示器可以显示各种提示、采集的人脸图像和/或各种界面。
本公开实施例还提供了一种计算机存储介质，计算机存储介质存储有计算机可执行代码；计算机可执行代码被执行后，能够实现前述一个或多个技术方案提供的图像处理方法，例如图1和/或图3所示的图像处理方法。
在本公开所提供的几个实施例中，应该理解到，所揭露的设备和方法，可以通过其它的方式实现。以上所描述的设备实施例仅仅是示意性的，例如，所述单元的划分，仅仅为一种逻辑功能划分，实际实现时可以有另外的划分方式，如：多个单元或组件可以结合，或可以集成到另一个系统，或一些特征可以忽略，或不执行。另外，所显示或讨论的各组成部分相互之间的耦合、或直接耦合、或通信连接可以是通过一些接口，设备或单元的间接耦合或通信连接，可以是电性的、机械的或其它形式的。
上述作为分离部件说明的单元可以是、或也可以不是物理上分开的，作为单元显示的部件可以是、或也可以不是物理单元，即可以位于一个地方，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全部单元来实现本实施例方案的目的。
另外，在本公开各实施例中的各功能单元可以全部集成在一个处理模块中，也可以是各单元分别单独作为一个单元，也可以两个或两个以上单元集成在一个单元中；上述集成的单元既可以采用硬件的形式实现，也可以采用硬件加软件功能单元的形式实现。
本公开任意实施例公开的技术特征，在不冲突的情况下，可以任意组合形成新的方法实施例或设备实施例。
本公开任意实施例公开的方法实施例，在不冲突的情况下，可以任意组合形成新的方法实施例。
本公开任意实施例公开的设备实施例，在不冲突的情况下，可以任意组合形成新的设备实施例。
本领域普通技术人员可以理解：实现上述方法实施例的全部或部分步骤可以通过程序指令相关的硬件来完成，前述的程序可以存储于一计算机可读取存储介质中，该程序在执行时，执行包括上述方法实施例的步骤；而前述的存储介质包括：移动存储设备、只读存储器(ROM，Read-Only Memory)、随机存取存储器(RAM，Random Access Memory)、磁碟或者光盘等各种可以存储程序代码的介质。
以上所述，仅为本公开的具体实施方式，但本公开的保护范围并不局限于此，任何熟悉本技术领域的技术人员在本申请揭露的技术范围内，可轻易想到变化或替换，都应涵盖在本公开的保护范围之内。因此，本公开的保护范围应以所述权利要求的保护范围为准。

Claims (18)

  1. 一种图像处理方法，所述方法包括：
    获取第一图像中目标对象包含的第一部位的第一关键点的位置信息；
    基于所述第一关键点的位置信息，确定包含所述第一关键点的第一区域；
    根据所述第一区域内预设的变形网格的网格点与所述第一区域内的像素点之间的相对位置，确定所述第一区域内网格点的位移量；
    根据所述第一区域内网格点的位移量，控制所述第一区域内像素点的变形，得到变形后的第二图像。
  2. 根据权利要求1所述的方法，其中，所述根据所述第一区域内预设的变形网格的网格点与所述第一区域内的像素点之间的相对位置，确定所述第一区域内网格点的位移量，包括：
    根据所述第一区域内的预设的变形网格的网格点与所述第一区域内的像素点之间的相对位置，确定所述网格点的位移量的衰减参数；以及，根据变形指令，确定所述网格点的第一位移量；
    根据所述衰减参数对所述第一位移量进行衰减处理，得到比所述第一位移量小的第二位移量。
  3. 根据权利要求2所述的方法，其中，所述根据所述第一区域内网格点的位移量，控制所述第一区域内像素点的变形，得到变形后的第二图像，包括：
    根据所述第二位移量，控制所述第一区域内的相邻像素点之间的间距，得到变形后的第二图像。
  4. 根据权利要求2所述的方法，其中，所述根据所述第一区域内预设的变形网格的网格点与所述第一区域内像素之间的相对位置，确定所述网格点的位移量的衰减参数，包括：
    确定位于多个所述第一关键点的连线上的像素点，得到第一集合；
    根据所述第一集合内像素点与所述变形网格内网格点的相对位置，得到第二集合，其中，所述第二集合包括：所述变形网格中与所述第一集合内各像素点分别距离最近的目标网格点；
    根据所述第二集合中各所述目标网格点与所述第一集合中由所述目标网格点所控制的像素点之间的相对位置，确定所述第二集合中各所述目标网格点的所述衰减参数。
  5. 根据权利要求4所述的方法，其中，所述根据所述第二集合中各所述目标网格点与所述第一集合中由所述目标网格点所控制的像素点之间的相对位置，确定所述第二集合中各所述网格点的所述衰减参数，包括：
    分别以各所述第一关键点为中心，在预定方向上向外进行遍历所述第二集合中各所述目标网格点，得到所述第二集合中各所述目标网格点相对于每个所述第一关键点在预定方向上的间距排序；
    根据所述间距排序，确定所述第二集合中各所述目标网格点的所述衰减参数。
  6. 根据权利要求5所述的方法，其中，所述根据所述间距排序，确定所述第二集合中各所述目标网格点的所述衰减参数，包括：
    在所述第二集合中任意一个所述目标网格点位于多个所述第一关键点的预定方向的情况下，根据多个所述第一关键点所对应的所述间距排序，确定所述衰减参数的备选值；
    选择所述备选值中的最大值，作为任意一个所述目标网格点的所述衰减参数。
  7. 根据权利要求1至6任一项所述的方法，其中，所述第一部位为上肢；所述获取第一图像中目标对象包含的第一部位的第一关键点的位置信息，包括：
    获取所述第一图像中上肢的骨架关键点的位置信息，所述骨架关键点包括以下关键点中的至少一种：肩部关键点、肘关节关键点、手腕关键点和手部关键点。
  8. 根据权利要求1至7任一项所述的方法，其中，所述方法还包括：
    根据所述第一图像中目标所包含的第二部位的第二关键点的位置信息，确定第二部位对应的第二区域；
    根据所述第二区域内的预设的变形网格的网格点的第一位移量，控制所述第一图像内第二区域的变形，得到变形后的第二图像。
  9. 一种图像处理装置，所述装置包括：
    获取模块，配置为获取第一图像中目标对象包含的第一部位的第一关键点的位置信息；
    第一确定模块，配置为基于所述第一关键点的位置信息，确定包含所述第一关键点的第一区域；
    第二确定模块，配置为根据所述第一区域内预设的变形网格的网格点与所述第一区域内的像素点之间的相对位置，确定所述第一区域内网格点的位移量；
    控制模块，配置为根据所述第一区域内网格点的位移量，控制所述第一区域内像素 点的变形，得到变形后的第二图像。
  10. 根据权利要求9所述的装置，其中，所述第二确定模块，配置为根据所述第一区域内的预设的变形网格的网格点与所述第一区域内的像素点之间的相对位置，确定所述网格点的位移量的衰减参数；以及，根据变形指令，确定所述网格点的第一位移量；根据所述衰减参数对所述第一位移量进行衰减处理，得到比所述第一位移量小的第二位移量。
  11. 根据权利要求10所述的装置，其中，所述控制模块，配置为根据所述第二位移量，控制所述第一区域内的相邻像素点之间的间距，得到变形后的第二图像。
  12. 根据权利要求10所述的装置，其中，所述第二确定模块，配置为确定位于多个所述第一关键点的连线上的像素点，得到第一集合；根据所述第一集合内像素点与所述变形网格内网格点的相对位置，得到第二集合，其中，所述第二集合包括：所述变形网格中与所述第一集合内各像素点距离最近的目标网格点；根据所述第二集合中各所述目标网格点与所述第一集合中由所述目标网格点所控制的像素点之间的相对位置，确定所述第二集合中各所述目标网格点的所述衰减参数。
  13. 根据权利要求12所述的装置，其中，所述第二确定模块，配置为分别以各所述第一关键点为中心，在预定方向上向外进行遍历所述第二集合中各所述目标网格点，得到所述第二集合中各所述目标网格点相对于每个所述第一关键点在预定方向上的间距排序；根据所述间距排序，确定所述第二集合中各所述目标网格点的所述衰减参数。
  14. 根据权利要求13所述的装置，其中，所述第二确定模块，配置为在所述第二集合中任意一个所述目标网格点位于多个所述第一关键点的预定方向的情况下，根据多个所述第一关键点所对应的所述间距排序，确定所述衰减参数的备选值；选择所述备选值中的最大值，作为任意一个所述目标网格点的所述衰减参数。
  15. 根据权利要求9至14任一项所述的装置，其中，所述第一部位为上肢；
    所述获取模块，配置为获取所述第一图像中上肢的骨架关键点的位置信息，所述骨架关键点包括以下关键点中的至少一种：肩部关键点、肘关节关键点、手腕关键点和手部关键点。
  16. 根据权利要求9至15任一项所述的装置，其中，所述第一确定模块，还配置为根据所述第一图像中目标所包含的第二部位的第二关键点的位置信息，确定第二部位对应的第二区域；
    所述控制模块，还配置为根据所述第二区域内的预设的变形网格的网格点的第一位 移量，控制所述第一图像内第二区域的变形，得到变形后的第二图像。
  17. 一种图像处理设备，包括：
    存储器；
    处理器，与所述存储器连接，用于通过执行存储在所述存储器上的计算机可执行指令实现权利要求1至8任一项提供的方法。
  18. 一种计算机存储介质，所述计算机存储介质存储有计算机可执行指令；所述计算机可执行指令被处理器执行后，能够实现权利要求1至8任一项提供的方法。
PCT/CN2020/093442 2019-12-25 2020-05-29 图像处理方法及装置、图像处理设备及存储介质 WO2021128731A1 (zh)

Priority Applications (4)

Application Number Priority Date Filing Date Title
KR1020207037116A KR20210084348A (ko) 2019-12-25 2020-05-29 이미지 처리 방법 및 장치, 이미지 처리 기기 및 저장 매체
SG11202109179WA SG11202109179WA (en) 2019-12-25 2020-05-29 Method and device for processing image, and storage medium
JP2020572416A JP7160958B2 (ja) 2019-12-25 2020-05-29 画像処理方法及び装置、画像処理デバイス並びに記憶媒体
US17/377,444 US20210342970A1 (en) 2019-12-25 2021-07-16 Method and device for processing image, and storage medium

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201911360894.4 2019-12-25
CN201911360894.4A CN111145084B (zh) 2019-12-25 2019-12-25 图像处理方法及装置、图像处理设备及存储介质

Related Child Applications (1)

Application Number Title Priority Date Filing Date
US17/377,444 Continuation US20210342970A1 (en) 2019-12-25 2021-07-16 Method and device for processing image, and storage medium

Publications (1)

Publication Number Publication Date
WO2021128731A1 true WO2021128731A1 (zh) 2021-07-01

Family

ID=70520246

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2020/093442 WO2021128731A1 (zh) 2019-12-25 2020-05-29 图像处理方法及装置、图像处理设备及存储介质

Country Status (7)

Country Link
US (1) US20210342970A1 (zh)
JP (1) JP7160958B2 (zh)
KR (1) KR20210084348A (zh)
CN (1) CN111145084B (zh)
SG (1) SG11202109179WA (zh)
TW (1) TW202125402A (zh)
WO (1) WO2021128731A1 (zh)

Families Citing this family (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111145084B (zh) * 2019-12-25 2023-06-16 北京市商汤科技开发有限公司 图像处理方法及装置、图像处理设备及存储介质
CN112767288B (zh) * 2021-03-19 2023-05-12 北京市商汤科技开发有限公司 图像处理方法及装置、电子设备和存储介质

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20050206739A1 (en) * 2004-03-19 2005-09-22 Yukio Kamoshida Image deformation estimating method and image deformation estimating apparatus
CN102971769A (zh) * 2010-04-30 2013-03-13 欧姆龙株式会社 图像变形装置、电子设备、图像变形方法、图像变形程序、以及记录有该程序的记录介质
CN110060287A (zh) * 2019-04-26 2019-07-26 北京迈格威科技有限公司 人脸图像鼻部整形方法及装置
CN110766607A (zh) * 2018-07-25 2020-02-07 北京市商汤科技开发有限公司 一种图像处理方法、装置和计算机存储介质
CN111145084A (zh) * 2019-12-25 2020-05-12 北京市商汤科技开发有限公司 图像处理方法及装置、图像处理设备及存储介质

Family Cites Families (9)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JPH0628465A (ja) * 1992-07-07 1994-02-04 Seiko Epson Corp 画像変換装置
US7952595B2 (en) * 2007-02-13 2011-05-31 Technische Universität München Image deformation using physical models
JP4289415B2 (ja) 2007-03-27 2009-07-01 セイコーエプソン株式会社 画像変形のための画像処理
CN102496140B (zh) * 2011-12-06 2013-07-31 中国科学院自动化研究所 一种基于多层嵌套笼体的实时交互式图像变形方法
JP6328669B2 (ja) 2013-02-23 2018-05-23 クゥアルコム・インコーポレイテッドＱｕａｌｃｏｍｍ Ｉｎｃｏｒｐｏｒａｔｅｄ 電子デバイスによる対話型画像戯画化のためのシステムおよび方法
TW201510771A (zh) * 2013-09-05 2015-03-16 Utechzone Co Ltd 指向位置偵測裝置及其方法、程式及電腦可讀取紀錄媒體
CN103824253B (zh) * 2014-02-19 2017-01-18 中山大学 一种基于图像局部精确变形的人物五官变形方法
CN110060348B (zh) * 2019-04-26 2023-08-11 北京迈格威科技有限公司 人脸图像整形方法及装置
CN110555798B (zh) * 2019-08-26 2023-10-17 北京字节跳动网络技术有限公司 图像变形方法、装置、电子设备及计算机可读存储介质

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20050206739A1 (en) * 2004-03-19 2005-09-22 Yukio Kamoshida Image deformation estimating method and image deformation estimating apparatus
CN102971769A (zh) * 2010-04-30 2013-03-13 欧姆龙株式会社 图像变形装置、电子设备、图像变形方法、图像变形程序、以及记录有该程序的记录介质
CN110766607A (zh) * 2018-07-25 2020-02-07 北京市商汤科技开发有限公司 一种图像处理方法、装置和计算机存储介质
CN110060287A (zh) * 2019-04-26 2019-07-26 北京迈格威科技有限公司 人脸图像鼻部整形方法及装置
CN111145084A (zh) * 2019-12-25 2020-05-12 北京市商汤科技开发有限公司 图像处理方法及装置、图像处理设备及存储介质

Also Published As

Publication number Publication date
US20210342970A1 (en) 2021-11-04
CN111145084B (zh) 2023-06-16
SG11202109179WA (en) 2021-09-29
KR20210084348A (ko) 2021-07-07
JP7160958B2 (ja) 2022-10-25
CN111145084A (zh) 2020-05-12
JP2022518641A (ja) 2022-03-16
TW202125402A (zh) 2021-07-01

Similar Documents

Publication Publication Date Title
TWI701941B (zh) 圖像處理方法及裝置、電子設備及儲存介質
WO2021128731A1 (zh) 图像处理方法及装置、图像处理设备及存储介质
WO2023273093A1 (zh) 一种人体三维模型获取方法、装置、智能终端及存储介质
TWI755768B (zh) 圖像處理方法、圖像處理設備及儲存介質
CN109961523B (zh) 虚拟目标的更新方法、装置、系统、设备及存储介质
CN113129450A (zh) 虚拟试衣方法、装置、电子设备和介质
US10147240B2 (en) Product image processing method, and apparatus and system thereof
CN113724391A (zh) 三维模型构建方法、装置、电子设备和计算机可读介质
CN112657176A (zh) 一种结合人像行为信息的双目投影人机交互方法
WO2022174574A1 (zh) 基于传感器的裸手数据标注方法及系统
CN107154028A (zh) 图像降噪方法及装置
CN115619986B (zh) 场景漫游方法、装置、设备和介质
JP2018132319A (ja) 情報処理装置、及び、情報処理装置の制御方法、コンピュータプログラム、記憶媒体
CN108256477B (zh) 一种用于检测人脸的方法和装置
CN116912387A (zh) 纹理贴图的处理方法及装置、电子设备、存储介质
CN110852943A (zh) 图像处理方法及装置、图像处理设备及存储介质
CN110197228B (zh) 图像校正方法及装置
CN108961389B (zh) 消除房源虚拟三维模型中镜面产生的虚像的方法及系统
JP7429666B2 (ja) 点群ノイズ除去装置及びプログラム
US20210042947A1 (en) Method and apparatus for processing data, electronic device and storage medium
CN116468829A (zh) 动画生成方法、装置、存储介质及电子设备
CN116958406A (zh) 一种人脸三维重建方法、装置、电子设备及存储介质
CN117275089A (zh) 一种单目摄像头的人物识别方法、装置、设备及存储介质
CN116958338A (zh) 对象姿态的调整方法、装置、设备、介质及产品
CN116416361A (zh) 图像生成方法、装置、电子设备及存储介质

Legal Events

Date Code Title Description
ENP Entry into the national phase

Ref document number: 2020572416

Country of ref document: JP

Kind code of ref document: A

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

Ref document number: 20905581

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: 20905581

Country of ref document: EP

Kind code of ref document: A1