+

WO2016175675A1 - Procédé de modification d'une image portrait et dispositif permettant de mettre en oeuvre ce procédé - Google Patents

Procédé de modification d'une image portrait et dispositif permettant de mettre en oeuvre ce procédé Download PDF

Info

Publication number
WO2016175675A1
WO2016175675A1 PCT/RU2015/000279 RU2015000279W WO2016175675A1 WO 2016175675 A1 WO2016175675 A1 WO 2016175675A1 RU 2015000279 W RU2015000279 W RU 2015000279W WO 2016175675 A1 WO2016175675 A1 WO 2016175675A1
Authority
WO
WIPO (PCT)
Prior art keywords
portrait image
template
image
portrait
source
Prior art date
Application number
PCT/RU2015/000279
Other languages
English (en)
Inventor
Fedor Valeryevich SAROKVASHA
Original Assignee
Sarokvasha Fedor Valeryevich
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 Sarokvasha Fedor Valeryevich filed Critical Sarokvasha Fedor Valeryevich
Priority to PCT/RU2015/000279 priority Critical patent/WO2016175675A1/fr
Publication of WO2016175675A1 publication Critical patent/WO2016175675A1/fr

Links

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V10/00—Arrangements for image or video recognition or understanding
    • G06V10/98—Detection or correction of errors, e.g. by rescanning the pattern or by human intervention; Evaluation of the quality of the acquired patterns
    • 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
    • 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
    • G06V40/16—Human faces, e.g. facial parts, sketches or expressions
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
    • H04N5/00—Details of television systems
    • H04N5/44—Receiver circuitry for the reception of television signals according to analogue transmission standards
    • H04N5/57—Control of contrast or brightness

Definitions

  • the present invention generally relates to the field of image processing, and more particularly to a method of modifying a portrait digital image and to a device for implementation such method.
  • raster graphic editors such as Adobe PhotoshopTM
  • images e.g. portrait images
  • These editors are multifunctional and may be used to modify a portrait image by overlying thereof with another portrait image.
  • These editors further may be used to provide color correction of obtained portrait image to compensate color difference between original image and overlaid image.
  • a person who is highly skilled and knowledgeable in the use of such editors is required to provide a combined portrait image that does not appear to be a result of combination and looks natural, so that creation of a mixed portrait image using at least two initial portrait images become cost consuming procedure. Therefore, proposed are various technical solutions aiming for simplification of modifying portrait images.
  • Known from EP 1376475 are an apparatus and method of modifying a portrait image.
  • the method comprises providing a digital still image comprising the portrait image; generating an image model of the portrait image; providing a plurality of modified portrait images generated by applying a plurality of pre-defined modifications to the image model; displaying the plurality of modified portrait images and selecting at least one of the plurality of displayed modified portrait images by a user to further display, transfer, store, or print such portrait image.
  • known method and device allows selection and modification of one portrait image but fails to disclose combining or overlying several portrait images.
  • a method of correcting face image and makeup simulation method includes mixing a desirable face with an actual face of a user to approach the latter.
  • the actual face of a user is subjected to image processing using modifying technique like morphing programs or twinning programs.
  • image processing fails to include color correction of the portrait images to obtain authentic portrait image.
  • an object of the present invention is to provide convenient tool for modifying a portrait image that is free of disadvantages of the known solutions.
  • Another object of the present invention is to provide simple, fast, cost effective and user friendly method of modifying a portrait image.
  • Still another object of the present invention is to provide real time technique of modifying a portrait image that provides automatic color correction of a portrait image to be pasted to obtain naturally combined collage of initial and pasted photo without using professional image processing tools.
  • a computer implemented method for modifying a template portrait image.
  • the method comprises providing a template portrait image having at least one person's face, providing a source portrait image having a person's face, color correction of the source portrait image by means of pixel-by-pixel linear transformation to obtain transformed source image having color characteristics similar to those of the template portrait image, overlaying the template portrait image with transformed source portrait image, positioning the overlaid transformed source portrait image on the template portrait image to obtain naturally combined modified template portrait image comprising collage of template and source images.
  • a computer device for implementation the method.
  • the computer device comprising display means for displaying a template portrait image, an image capturing module adapted to capture at least a source portrait image, a memory for storing at least the template portrait images, a central processing unit adapted to receive the data from the display means, image capture module and the memory to process them in order to obtain a modified template portrait image comprising collage of the template portrait image and the source portrait images.
  • the term “computer device” refers to an electronic device comprising at least one central processing unit (CPU), an image capture module, a memory, display means, input means for input data or commands to the device, and means for data transmission to transmit data outside the device, e.g. to the local network (Intranet), global network (Internet) or to peripheral devices (printers, scanners or the like).
  • the computer device according to the present application may be cell phone, smartphone, pad computer. Other types of computer devices having the similar architecture and functionality may also be used.
  • the term “portrait image” is equivalent to the term “portrait”. Both these terms refer to a digital portrait image especially one showing the face.
  • the term “portrait image” or “portrait” encompasses an image comprising a representation of at least one face, including human/person or animal.
  • the term further comprises any other elements of the human/person elements such as hairs, hats, glasses or the like.
  • memory refers both to a memory of the computer device and the remote access memory storages also referred to as Cloud storages.
  • template portrait image or “template portrait” refers to a digital image having at least one person's face represented thereon.
  • the template portrait images are subjects of modification according to the claimed invention.
  • Each template portrait image is stored in the memory along with corresponding accompanying data.
  • the accompanying data are obtained as a result of preliminary processing of each template portrait image.
  • the template portrait image may be an arbitrary portrait image provided by a user, e.g. using image capture module, that is not preliminary processed.
  • the accompanying data in this case are obtained during implementation of the claimed method.
  • source portrait image refers to a digital image provided by a user and having a person's face thereon.
  • a user may capture the source portrait image using an image capturing module (camera) of the computer device or the source portrait image may be selected from gallery of images stored in the memory.
  • the template portrait and source portrait are preferably images of the same size (640px x 640 px), so that they are able to be placed and processed in common coordinate system.
  • transformed source portrait image or “transformed source portrait” refers to a digital image achieved by results of pixel-by-pixel transformation of HSB (Hue, Saturation, Brightness) values of each pixel of the source portrait image.
  • the transformed source portrait image is used for further overlaying the template portrait image to form naturally combined collage of template and source images, which collage may be displayed, transferred, stored or printed.
  • image point refers to one pixel of a portrait image.
  • a smartphone such as one of those commercially available smartphones with touchscreen, is selected as a computer device for implementation of the claimed method.
  • the smartphone comprises at least a CPU, a memory, a digital camera, a touch screen display and a communication module.
  • the template portrait image is overlaid with the transformed source portrait image provided by the user.
  • the template portrait image is initially provided, wherein the template portrait image is taken from the memory of the smartphone.
  • the accompanying data of the template portrait image include HSB values (color data) of at least one image point (reference point) as will be described in more details below.
  • the image point is an arbitrary image point taken in the template portrait image.
  • the image point is chosen on the face region, preferably the region free from facial defects such as birthmarks, pimples or the like, preferably in the center of frontal area of the face, so that the HSB values of the point are close to those of majority of other points on the template portrait image.
  • the values are represented in digital format from the range [0 - 1] and stored in the memory in correspondence with particular template portrait image.
  • the accompanying data further comprises spatial coordinates of person's left eye and right eye and may comprise spatial coordinate of the central point of the upper lip.
  • the digital data, such as HSB values and the coordinates are organized as text files and may be stored in the memory.
  • template portrait images It is preferably to have front-facing photos with good lighting conditions (with no shadows) as template portrait images.
  • the user provide a template portrait image by sending request to a database stored in the memory consisting of template portrait images each of which is accompanied with digital data such as HSB values of the reference point and eyes coordinates. Once the user provides a template portrait image, the next step of the method is initiated.
  • the user provides a source portrait image.
  • the user may capture his/her own photo or choose digital photo from personal photo gallery stored in the memory.
  • it is preferably to capture or choose front-facing photo with good lighting conditions, especially with incidence of light similar to that on the template portrait.
  • a user may see and adjust his/her position to capture a photo of best quality.
  • the user may be guided by tips displayed on the display of the smartphone during capturing a photo.
  • an oval or rectangular guiding frame may be provided on the display to assist the user in proper photo capturing.
  • Corresponding recommendations to enlarge or reduce an image may also be provided, if the source portrait image is the image chosen from the gallery stored in the memory. Once the allowable source portrait image is captured, spatial coordinates of the face eyes shown on the source portrait are obtained, the spatial coordinates are stored so as to be accessible during further steps of the method. Once the source portrait image is captured or chosen next step of the method is initiated.
  • an automatic color correction using smartphone hardware comprises pixel-by-pixel linear transformation of the source portrait image in order to smooth color difference between the template portrait image and the source portrait image.
  • a transformed source portrait image having modified (HSB)' values for each pixel similar to HSB values of the template portrait image are obtained.
  • the color correction makes it possible to further obtain naturally combined collage of template and source portraits.
  • the color correction is initiated automatically as soon as a user inputs command to paste the source portrait image to the template portrait image.
  • the template portrait image is preliminary processed, and the template portrait image is stored in the memory along with accompanying data.
  • the data comprise at least HSB values of at least one reference point of the template portrait image and eyes coordinates of the face shown thereon.
  • the HSB values according to the referred embodiment are obtained as follows. Taken is an array of image points (e.g. [3x3] array) positioned approximately in the forehead center of a person's shown on the template portrait image, wherein any image points falling on facial defects such as birthmarks, pimples or the like, are not considered. Further, HSB values of each image point are obtained. Then, average values (HSB)i av are calculated as follows:
  • the values (HSB)i av are stored as digital data along with the template portrait image as a text file.
  • the average (HSB)i av values of the template portrait image are compared with those of the source portrait image.
  • defined are spatial coordinates (xi, yi) and (xi, yi) of the person's eyes, preferably pupils, on the source portrait image.
  • a reference image point (rx, ry) is calculated as follows: [0026] Further, an image points array having at least [3x3] dimensions is formed, wherein the reference image point (rx, ry) is taken as a central point of the array.
  • the reference point is approximately in the forehead center of a person's face shown on the source portrait image.
  • calculation to obtain the average (HSB) ⁇ values of the source portrait image is provided in a manner as described above:
  • dH -Hl at >— H2at>
  • dS Sia V — ⁇ 3 ⁇ 4 ⁇
  • B B ⁇ mv — B ⁇ av wherein dH, dS, dB - Hue, Saturation and Brightness value difference between average values HSBi av and HSB2av
  • the obtained (dH, dS, dB) values are further used for pixel-be-pixel linear transformation of the source portrait image to obtain transformed source portrait image.
  • the transformation is provided by assigning new H'S'B values to each pixel of the source portrait image based on the following rule:
  • the template portrait image is overlaid with the transformed source portrait image, that is positioned to fit the template portrait image and to form naturally combined collage of the template and source portraits.
  • the obtained values dx, dy define coordinate shift of the transformed source portrait with respect to the template portrait to match the left eyes on these two images.
  • dpix, dpiy, dp2X, dpiy are obtained as follows:
  • the method further comprises overlying the portrait image obtained at the previous stage with a mask.
  • the mask is a digital image obtained by cutting the face to be replaced from the template portrait image.
  • the additionally overlaid image is one more digital image obtained in advance by processing the template portrait image, e.g. when HSB values and eyes coordinates are determined.
  • the mask may be a file of the same format as the template portrait image and may be stored in the memory, so that once a user selects a template portrait image to be modified, corresponding mask is enabled.
  • the mask may be obtained by known graphic editors and the user need not make any operations with the mask during implementation of the method.
  • the use of mask helps to achieve even more natural and authentic portrait photo, e.g. when a face of the template portrait belongs to a woman and the face of the source portrait belongs to a men.
  • the modified portrait image is displayed on the screen of the smartphone as soon as the method is implemented and may be stored in the memory or transferred via Internet, e.g. shared vie social networks such as Facebook or Instagram

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Multimedia (AREA)
  • Theoretical Computer Science (AREA)
  • Quality & Reliability (AREA)
  • Processing Or Creating Images (AREA)

Abstract

L'invention concerne un dispositif et un procédé de modification d'une image numérique de portrait. Ledit procédé et ledit dispositif fournissent des outils automatiques simples pour modifier une image portrait en particulier en appliquant une correction de couleurs à la zone faciale superposée de manière à obtenir une superposition combinée naturellement de l'image portrait initiale et de l'image portrait source sans outils de traitement d'image professionnel.
PCT/RU2015/000279 2015-04-29 2015-04-29 Procédé de modification d'une image portrait et dispositif permettant de mettre en oeuvre ce procédé WO2016175675A1 (fr)

Priority Applications (1)

Application Number Priority Date Filing Date Title
PCT/RU2015/000279 WO2016175675A1 (fr) 2015-04-29 2015-04-29 Procédé de modification d'une image portrait et dispositif permettant de mettre en oeuvre ce procédé

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
PCT/RU2015/000279 WO2016175675A1 (fr) 2015-04-29 2015-04-29 Procédé de modification d'une image portrait et dispositif permettant de mettre en oeuvre ce procédé

Publications (1)

Publication Number Publication Date
WO2016175675A1 true WO2016175675A1 (fr) 2016-11-03

Family

ID=57198546

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/RU2015/000279 WO2016175675A1 (fr) 2015-04-29 2015-04-29 Procédé de modification d'une image portrait et dispositif permettant de mettre en oeuvre ce procédé

Country Status (1)

Country Link
WO (1) WO2016175675A1 (fr)

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20030234871A1 (en) * 2002-06-25 2003-12-25 Squilla John R. Apparatus and method of modifying a portrait image
US20040085324A1 (en) * 2002-10-25 2004-05-06 Reallusion Inc. Image-adjusting system and method
US20060132506A1 (en) * 1997-03-06 2006-06-22 Ryuichi Utsugi Method of modifying facial images, makeup simulation method, makeup method, makeup support apparatus and foundation transfer film
RU2418683C2 (ru) * 2006-08-18 2011-05-20 Гемальто Ой Способ наложения одного изображения на другое, способ персонализации носителя данных с использованием способа наложения изображений и персонализированный носитель данных

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20060132506A1 (en) * 1997-03-06 2006-06-22 Ryuichi Utsugi Method of modifying facial images, makeup simulation method, makeup method, makeup support apparatus and foundation transfer film
US20030234871A1 (en) * 2002-06-25 2003-12-25 Squilla John R. Apparatus and method of modifying a portrait image
US20040085324A1 (en) * 2002-10-25 2004-05-06 Reallusion Inc. Image-adjusting system and method
RU2418683C2 (ru) * 2006-08-18 2011-05-20 Гемальто Ой Способ наложения одного изображения на другое, способ персонализации носителя данных с использованием способа наложения изображений и персонализированный носитель данных

Similar Documents

Publication Publication Date Title
US12223541B2 (en) Virtual try-on system, virtual try-on method, computer program product, and information processing device
KR102728245B1 (ko) 아바타 표시 장치, 아바타 생성 장치 및 프로그램
US12254565B2 (en) Visualization of post-treatment outcomes for medical treatment
JP6864449B2 (ja) イメージの明るさを調整する方法及び装置
WO2021169307A1 (fr) Procédé et appareil de traitement d'essayage de maquillage pour image de visage, dispositif informatique et support d'enregistrement
CN111784568A (zh) 人脸图像处理方法、装置、电子设备及计算机可读介质
CN110738732A (zh) 一种三维人脸模型的生成方法和设备
CN111882627A (zh) 图像处理方法、视频处理方法、装置、设备和存储介质
CN111263066A (zh) 构图指导方法、装置、电子设备及存储介质
US20200394828A1 (en) Image object pose synchronization
CN104811684A (zh) 一种图像的三维美颜方法及装置
CN107578371A (zh) 一种图像处理方法、装置、电子设备及介质
CN110503599B (zh) 图像处理方法和装置
CN103198469B (zh) 图像处理装置以及图像处理方法
US9020255B2 (en) Image processing apparatus, image processing method, and storage medium
CN111654624B (zh) 拍摄提示方法、装置及电子设备
CN108227925A (zh) 一种坐姿调整的方法、装置、设备及存储介质
JP2019046239A (ja) 画像処理装置、画像処理方法、プログラム及び合成用画像データ
CN110610469A (zh) 一种人脸图像隐私保护方法、装置、设备及存储介质
KR102275172B1 (ko) 술통형 왜곡 보정 및 배경교체 기능 기반 셀프 증명사진 촬영이 가능한 사진 촬영 서비스 제공 방법
US10354125B2 (en) Photograph processing method and system
WO2016175675A1 (fr) Procédé de modification d'une image portrait et dispositif permettant de mettre en oeuvre ce procédé
JP6650738B2 (ja) 情報処理装置、情報処理システム、情報処理方法及びプログラム
JP7086362B1 (ja) 情報処理システム、情報処理方法および情報処理プログラム
WO2022205167A1 (fr) Procédé et appareil de traitement d'image, plateforme mobile, dispositif terminal et support de stockage

Legal Events

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

Ref document number: 15890868

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: 15890868

Country of ref document: EP

Kind code of ref document: A1

点击 这是indexloc提供的php浏览器服务，不要输入任何密码和下载