WO2016098943A1 - Procédé et système de traitement d'images pour améliorer la capacité de détection de visages - Google Patents

Procédé et système de traitement d'images pour améliorer la capacité de détection de visages Download PDF

Info

Publication number
WO2016098943A1
WO2016098943A1 PCT/KR2015/001366 KR2015001366W WO2016098943A1 WO 2016098943 A1 WO2016098943 A1 WO 2016098943A1 KR 2015001366 W KR2015001366 W KR 2015001366W WO 2016098943 A1 WO2016098943 A1 WO 2016098943A1
Authority
WO
WIPO (PCT)
Prior art keywords
image
histogram
unsharp masking
original image
image processing
Prior art date
Application number
PCT/KR2015/001366
Other languages
English (en)
Korean (ko)
Inventor
김태성
경종민
황영배
Original Assignee
재단법인 다차원 스마트 아이티 융합시스템 연구단
전자부품연구원
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 재단법인 다차원 스마트 아이티 융합시스템 연구단, 전자부품연구원 filed Critical 재단법인 다차원 스마트 아이티 융합시스템 연구단
Publication of WO2016098943A1 publication Critical patent/WO2016098943A1/fr

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T5/00—Image enhancement or restoration
    • G06T5/40—Image enhancement or restoration using histogram techniques
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis

Definitions

  • the present invention relates to an image processing system and method for improving face detection, and more particularly, to processing an image by performing histogram equalization and selectively applying unsharp masking. It's about technology.
  • Existing image processing techniques process an image by performing histogram flattening or histogram specification on all of a plurality of pixels included in the original image, or by applying unsharp masking to the original image unconditionally.
  • the present specification proposes an image processing technique that considers a saturated region among a plurality of pixels included in an original image.
  • Embodiments of the present invention provide an image processing method and system considering a saturated region among a plurality of pixels included in an original image.
  • embodiments of the present invention perform histogram flattening on at least some of the plurality of pixels included in the original image, and unsharp mask the image on which the histogram flattening is performed based on the histogram extracted from the original image.
  • an image processing method and a system for selectively applying are provided.
  • An image processing method for improving face detection capability includes performing histogram equalization on at least some of a plurality of pixels included in an original image; And selectively applying unsharp masking to the image on which the histogram flattening is performed, based on the histogram extracted from the original image.
  • the step of performing histogram planarization may be a step of performing histogram planarization on at least some of the pixels other than a saturation region of the plurality of pixels included in the original image.
  • Selectively applying the unsharp masking may include calculating a ratio of saturated areas among a plurality of pixels included in the original image based on a histogram extracted from the original image; And applying unsharp masking to the image on which the histogram planarization is performed when the ratio of the calculated saturation region is equal to or greater than a preset reference value.
  • Selectively applying the unsharp masking may include blurring an image on which the histogram planarization has been performed; Inverting the blurred image; And combining the inverted image and the image on which the histogram planarization has been performed.
  • Inverting the blurred image further includes scaling the inverted image, and combining the histogram flattened image and the inverted image comprises: Combining the performed image with the scaled image.
  • the performing of the histogram planarization may further include extracting a histogram for a plurality of pixels included in the original image from the original image.
  • the image processing method may further include performing face detection based on an image to which the unsharp masking is selectively applied.
  • An image processing system for improving face detection capability includes a histogram flattening unit performing histogram equalization on at least some of a plurality of pixels included in an original image; And an unsharp masking unit selectively applying unsharp masking to the image on which the histogram flattening is performed, based on the histogram extracted from the original image.
  • the histogram planarizer may perform histogram planarization on at least some of the pixels other than a saturation region of the plurality of pixels included in the original image.
  • the unsharp masking unit calculates a ratio of a saturation region among a plurality of pixels included in the original image based on the histogram extracted from the original image, and when the ratio of the calculated saturation region is greater than or equal to a preset reference value, the histogram Unsharp masking may be applied to an image on which planarization has been performed.
  • the unsharp masking unit may include: a blurring unit for blurring the image on which the histogram planarization is performed; An inverting unit for inverting the blurred image; And an image combiner configured to combine the image on which the histogram planarization is performed and the inverted image.
  • the inverting unit may further include a scale unit for scaling the inverted image, and the image combiner may combine the scaled image and the image on which the histogram planarization is performed.
  • the histogram flattener may further include a histogram extractor that extracts a histogram of a plurality of pixels included in the original image from the original image.
  • the image processing system may further include a face detector configured to perform face detection based on an image to which the unsharp masking is selectively applied.
  • Embodiments of the present invention may provide an image processing method and system that considers a saturated region among a plurality of pixels included in an original image.
  • embodiments of the present invention perform histogram flattening on at least some of the plurality of pixels included in the original image, and unsharp mask the image on which the histogram flattening is performed based on the histogram extracted from the original image. It is possible to provide an image processing method and system for selectively applying.
  • FIG. 1 is a view showing an image processing method according to an embodiment of the present invention.
  • FIG. 2 is a diagram illustrating a histogram planarization performed according to an embodiment of the present invention.
  • FIG. 3 is a view for explaining the unsharp masking applied according to an embodiment of the present invention.
  • FIG. 4 is a view for explaining face detection performed according to an embodiment of the present invention.
  • FIG. 5 is a block diagram illustrating an image processing system according to an exemplary embodiment of the present invention.
  • FIG. 1 is a view showing an image processing method according to an embodiment of the present invention.
  • the image processing system performs a histogram flattening on the original image 110, and selectively applies unsharp masking to the image 120 on which the histogram flattening is performed.
  • An image processing method for improving detection capability is performed.
  • the image processing system first performs histogram planarization on at least some of the plurality of pixels included in the original image 110.
  • the image processing system performs histogram flattening on at least some pixels of the plurality of pixels included in the original image 110 except for the saturated region.
  • the image processing system may perform histogram planarization on at least some of the pixels except for the saturation region of the pixels having the brightness value of 255 among the plurality of pixels included in the original image 110.
  • the histogram flattening is a technique for performing contrast adjustment in the histogram, and a detailed description thereof will be omitted since it departs from the technical idea of the present invention.
  • the image processing system then selectively applies unsharp masking to the image 120 on which image flattening has been performed, based on the histogram extracted from the original image 110. For example, the image processing system calculates the ratio of the saturated areas among the plurality of pixels included in the original image 110 based on the histogram extracted from the original image 110, and then calculates the ratio of the calculated saturated areas in advance. If the threshold value is greater than or equal to the predetermined reference value, the unsharp masking may be applied to the image 120 on which the histogram planarization is performed. As a more specific example, the image processing system may apply unsharp masking to the image 120 on which the histogram planarization is performed when the ratio of the saturated areas among the plurality of pixels is 30% or more.
  • the image processing system selectively applies the unsharp masking by extracting a histogram for the plurality of pixels included in the original image 110 from the original image 110 before selectively applying the unsharp masking. It may be used in the process (for example, calculating the ratio of the saturation region of the plurality of pixels included in the original image 110).
  • the image processing system may not apply unsharp masking to the image 120 on which the histogram planarization is performed when the ratio of the calculated saturation region is less than a preset reference value.
  • the image processing system may not apply unsharp masking to the image 120 on which the histogram planarization is performed when the ratio of saturated regions among the plurality of pixels is less than 30%.
  • the image processing system blurs the image 120 on which the histogram flattening is performed, inverts the blurred image 130, and then the image 120 on which the histogram flattening is performed; By combining the inverted image 140, unsharp masking may be applied to the image 120 on which the histogram planarization has been performed. Accordingly, the image processing system may perform face detection based on the image 150 to which unsharp masking is selectively applied.
  • the image processing system may blur the image 120 on which the histogram planarization has been performed by using a blurring filter, as shown in Equation 1 below.
  • I_2 (x, y) F (I_1 (x, y))
  • I_2 (x, y) means a blurred image 130
  • F means a blurring filter (eg, an average filter or Gaussian filter) function
  • I_1 (x , y) means the image 120 on which the histogram planarization has been performed.
  • the image processing system may invert the blurred image 130, as shown in Equation 2.
  • I_3 (x, y) 255-I_2 (x, y)
  • I_3 (x, y) means the inverted image 140.
  • the image processing system may combine the inverted image 140 and the image 120 on which the histogram planarization is performed, as shown in Equation 3 below.
  • I_4 (x, y) I_1 (x, y) + I_3 (x, y)
  • I_4 (x, y) means the image 150 to which unsharp masking is selectively applied.
  • the image processing system may additionally perform scaling in the process of applying unsharp masking to the image 120 on which the histogram planarization has been performed. For example, the image processing system blurs the histogram flattened image 120, inverts the blurred image 130, and then scales the inverted image 140, thereby smoothing the histogram flattened.
  • the combined image 120 and the scaled image may be combined.
  • the image processing system may scale the inverted image 140, as shown in equation (4).
  • I_5 (x, y) I_3 (x, y) ⁇ 1 / (scale factor)
  • I_5 (x, y) means a scaled image
  • the scale factor may be adaptively changed to maximize edge enhancement of the scaled image
  • the image 150 to which the unsharp masking is selectively applied refers to an image in which the histogram flattening image 120 and the scaled image are combined, as shown in Equation 5 below.
  • I_4 (x, y) I_1 (x, y) + I_5 (x, y)
  • unsharp masking in which scaling is additionally performed may be adaptively performed when the ratio of the calculated saturation region described above is less than a predetermined reference value.
  • the image processing system may apply unsharp masking in which scaling is additionally performed if the calculated percentage of saturated areas is less than 30% and greater than 20%.
  • the image processing system may perform histogram flattening on all of the plurality of pixels included in the original image 110.
  • the saturation region of the plurality of pixels included in the original image 110 may be selectively frozen.
  • FIG. 2 is a diagram illustrating a histogram planarization performed according to an embodiment of the present invention.
  • the image processing system improves face detection capability by performing histogram planarization on at least some pixels other than a saturated region among a plurality of pixels included in an original image. Can be.
  • the vertical axis means population.
  • the horizontal axis means the brightness value
  • the face detection capability in the histogram flattening image 220 may be improved.
  • the face detection rate in the original image 210 is 84%
  • the face detection rate in the image 220 on which the histogram planarization is performed may be 91%.
  • FIG. 3 is a view for explaining the unsharp masking applied according to an embodiment of the present invention.
  • an image processing system selectively applies unsharp masking to an image 310 on which histogram flattening is performed, based on a histogram extracted from an original image, thereby detecting a face. Can improve.
  • edge emphasis graphs 311 and 321 in the image 310 to which the histogram flattening is performed and the image sharpening to the unsharp masking 320 in which the blurring, inverting, and image combining are sequentially performed edge emphasis
  • the vertical axis means pixel brightness value and the horizontal axis means pixel sequence
  • the face detection rate in the image 320 to which the unsharp masking is performed is improved.
  • the image processing system calculates the ratio of saturated areas among the plurality of pixels included in the original image based on the histogram extracted from the original image, and selectively applies such unsharp masking according to the calculated ratio of saturated areas.
  • the image processing may be performed considering the saturated region in the original image.
  • FIG. 4 is a view for explaining face detection performed according to an embodiment of the present invention.
  • an image processing system uses various face detection techniques based on an image to which unsharp masking is selectively applied (for example, may only be an image whose only histogram planarization has been performed). By doing this, face detection can be performed.
  • the image processing system may perform face detection by using a local binary pattern (LBP) technique based on an image to which unsharp masking is selectively applied.
  • LBP local binary pattern
  • the face detection rate is highest in (b) 410, and in the backlight, the original is detected.
  • the histogram planarization is performed on at least some pixels included in the image, and the unsharp masking is applied, it can be seen that the face detection rate is the highest at (e) 420.
  • the face detection rate may vary according to the ratio of saturated areas among the plurality of pixels included in the original image under various lighting environments.
  • the image processing system according to the embodiment of the present invention performs histogram flattening on at least some pixels in consideration of the ratio of saturated areas among the plurality of pixels included in the original image, and freezes the image on which the image flattening is performed.
  • the face detection capability can be adaptively improved under various lighting environments.
  • FIG. 5 is a block diagram illustrating an image processing system according to an exemplary embodiment of the present invention.
  • an image processing system includes a histogram flattening unit 510 and an unsharp masking unit 520.
  • the histogram flattener 510 performs histogram flattening on at least some of the plurality of pixels included in the original image.
  • the histogram flattener 510 performs histogram flattening on at least some pixels other than a saturated region among the plurality of pixels included in the original image.
  • the unsharp masking unit 520 selectively applies unsharp masking to the image on which image flattening is performed, based on the histogram extracted from the original image.
  • the unsharp masking unit 520 calculates a saturation region ratio among the plurality of pixels included in the original image based on the histogram extracted from the original image, and then calculates the ratio of the calculated saturation region to a preset reference value.
  • unsharp masking may be applied to the image on which the histogram planarization is performed.
  • the histogram flattening unit 510 further includes a histogram extracting unit for extracting a histogram for a plurality of pixels included in the original image from the original image before selectively applying unsharp masking. Accordingly, the histogram extracted from the original image may be used in the process of selectively applying unsharp masking (eg, calculating a ratio of saturated areas among a plurality of pixels included in the original image).
  • the unsharp masking unit 520 may not apply the unsharp masking to the image on which the histogram planarization is performed when the calculated ratio of the saturated areas is less than a predetermined reference value.
  • the unsharp masking unit 520 may include a blurring unit for blurring the image on which the histogram flattening is performed, an inverting unit for inverting the blurred image, and an image and inverting the histogram flattening.
  • a blurring unit for blurring the image on which the histogram flattening is performed
  • an inverting unit for inverting the blurred image
  • an image and inverting the histogram flattening By including an image combiner that combines the combined images, it is possible to apply unsharp masking to the image on which the histogram planarization has been performed.
  • the face detection unit included in the image processing system may perform face detection based on an image to which unsharp masking is selectively applied.
  • the unsharp masking unit 520 may additionally perform scaling in the process of applying unsharp masking to the image on which the histogram planarization is performed.
  • the unsharp masking unit 520 further includes a scale unit for scaling the inverted image, thereby blurring the image where the histogram flattening has been performed, inverting the blurred image, and additionally, inverting the inverted image.
  • an image combiner can be used to combine the scaled image and the image on which the histogram planarization has been performed.
  • the image processing system including the histogram flattening unit 510 and the unsharp masking unit 520 instead of performing histogram flattening on all the pixels included in the original image.
  • the saturation of the plurality of pixels included in the original image is applied.
  • by selectively applying unsharp masking it is possible to improve the face detection ability in an environment in which light changes are sensitive, such as backlighting.
  • the apparatus described above may be implemented as a hardware component, a software component, and / or a combination of hardware components and software components.
  • the devices and components described in the embodiments may be, for example, processors, controllers, arithmetic logic units (ALUs), digital signal processors, microcomputers, field programmable arrays (FPAs), It may be implemented using one or more general purpose or special purpose computers, such as a programmable logic unit (PLU), microprocessor, or any other device capable of executing and responding to instructions.
  • the processing device may execute an operating system (OS) and one or more software applications running on the operating system.
  • the processing device may also access, store, manipulate, process, and generate data in response to the execution of the software.
  • OS operating system
  • the processing device may also access, store, manipulate, process, and generate data in response to the execution of the software.
  • processing device includes a plurality of processing elements and / or a plurality of types of processing elements. It can be seen that it may include.
  • the processing device may include a plurality of processors or one processor and one controller.
  • other processing configurations are possible, such as parallel processors.
  • the software may include a computer program, code, instructions, or a combination of one or more of the above, and configure the processing device to operate as desired, or process it independently or collectively. You can command the device.
  • Software and / or data may be any type of machine, component, physical device, virtual equipment, computer storage medium or device in order to be interpreted by or to provide instructions or data to the processing device. Or may be permanently or temporarily embodied in a signal wave to be transmitted.
  • the software may be distributed over networked computer systems so that they may be stored or executed in a distributed manner.
  • Software and data may be stored on one or more computer readable recording media.
  • the method according to the embodiment may be embodied in the form of program instructions that can be executed by various computer means and recorded in a computer readable medium.
  • the computer readable medium may include program instructions, data files, data structures, etc. alone or in combination.
  • the program instructions recorded on the media may be those specially designed and constructed for the purposes of the embodiments, or they may be of the kind well-known and available to those having skill in the computer software arts.
  • Examples of computer-readable recording media include magnetic media such as hard disks, floppy disks, and magnetic tape, optical media such as CD-ROMs, DVDs, and magnetic disks, such as floppy disks.
  • Examples of program instructions include not only machine code generated by a compiler, but also high-level language code that can be executed by a computer using an interpreter or the like.
  • the hardware device described above may be configured to operate as one or more software modules to perform the operations of the embodiments, and vice versa.

Landscapes

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

Abstract

La présente invention concerne un procédé de traitement d'images permettant d'améliorer la capacité de détection de visages. Ledit procédé comprend les étapes de : réalisation d'une égalisation d'histogramme sur au moins quelques pixels d'une pluralité de pixels compris dans une image originale ; et application sélective d'un masque flou à une image sur laquelle l'égalisation d'histogramme a été réalisée, sur la base d'un histogramme extrait de l'image originale.
PCT/KR2015/001366 2014-12-18 2015-02-11 Procédé et système de traitement d'images pour améliorer la capacité de détection de visages WO2016098943A1 (fr)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
KR10-2014-0183555 2014-12-18
KR1020140183555A KR101617551B1 (ko) 2014-12-18 2014-12-18 얼굴 검출 능력 향상을 위한 이미지 처리 방법 및 시스템

Publications (1)

Publication Number Publication Date
WO2016098943A1 true WO2016098943A1 (fr) 2016-06-23

Family

ID=56022696

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/KR2015/001366 WO2016098943A1 (fr) 2014-12-18 2015-02-11 Procédé et système de traitement d'images pour améliorer la capacité de détection de visages

Country Status (2)

Country Link
KR (1) KR101617551B1 (fr)
WO (1) WO2016098943A1 (fr)

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111835954A (zh) * 2019-04-14 2020-10-27 泰州腾翔信息科技有限公司 现场数据抓取方法
CN114359030A (zh) * 2020-09-29 2022-04-15 合肥君正科技有限公司 一种人脸逆光图片的合成方法

Families Citing this family (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
KR102463017B1 (ko) * 2020-09-08 2022-11-03 주식회사 포스코 연원료 측량 시스템 및 그 측량 방법

Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2000285232A (ja) * 1999-03-31 2000-10-13 Seiko Epson Corp 画像鮮鋭化プログラムを記録した媒体、画像鮮鋭化装置および画像鮮鋭化方法
KR20010027675A (ko) * 1999-09-15 2001-04-06 윤종용 영상의 역광 보정장치
JP2003281534A (ja) * 2002-03-22 2003-10-03 Seiko Epson Corp 画像処理装置および画像出力装置

Patent Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2000285232A (ja) * 1999-03-31 2000-10-13 Seiko Epson Corp 画像鮮鋭化プログラムを記録した媒体、画像鮮鋭化装置および画像鮮鋭化方法
KR20010027675A (ko) * 1999-09-15 2001-04-06 윤종용 영상의 역광 보정장치
JP2003281534A (ja) * 2002-03-22 2003-10-03 Seiko Epson Corp 画像処理装置および画像出力装置

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
LEE, IM GEUN ET AL.: "Image Enhancement using Statistical Information of Pixel Dynamics", JOURNAL OF THE KOREAN INSTITUTE OF COMMUNICATION SCIENCES, vol. 12, no. 12, December 2008 (2008-12-01), pages 2337 - 2342 *

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111835954A (zh) * 2019-04-14 2020-10-27 泰州腾翔信息科技有限公司 现场数据抓取方法
CN114359030A (zh) * 2020-09-29 2022-04-15 合肥君正科技有限公司 一种人脸逆光图片的合成方法
CN114359030B (zh) * 2020-09-29 2024-05-03 合肥君正科技有限公司 一种人脸逆光图片的合成方法

Also Published As

Publication number Publication date
KR101617551B1 (ko) 2016-05-03

Similar Documents

Publication Publication Date Title
WO2021201438A1 (fr) Système et procédé de déformation de mouvement utilisant des trames à multiples expositions
WO2020231065A1 (fr) Procédé et appareil pour combiner des trames d'image capturées à l'aide de différents réglages d'exposition, en des images combinées
WO2021107610A1 (fr) Procédé et système de production d'une carte triple pour un matage d'image
WO2019156524A1 (fr) Appareil de traitement d'image, et procédé de traitement d'image associé
WO2016098943A1 (fr) Procédé et système de traitement d'images pour améliorer la capacité de détection de visages
WO2017014396A1 (fr) Dispositif et procédé pour débrumage d'image adaptatif à la longueur d'onde
WO2016089114A1 (fr) Procédé et appareil de floutage d'image
WO2020231035A1 (fr) Appareil de traitement d'image et son procédé de fonctionnement
WO2015182802A1 (fr) Procédé de traitement d'image et système l'utilisant, et support d'enregistrement pour son exécution
EP4172929A1 (fr) Définition des contours d'image
WO2016190698A1 (fr) Procédé de filtrage médian pondéré et dispositif de suppression de bruit
WO2023210884A1 (fr) Dispositif et procédé d'élimination de bruit basés sur de moyens non locaux
WO2020138630A1 (fr) Dispositif d'affichage et procédé de traitement d'image associé
WO2023068655A1 (fr) Système et procédé d'apprentissage de courbes de tons pour une amélioration d'image locale
WO2013077546A1 (fr) Appareil et procédé de détection d'une modification d'une scène dans une vidéo stéréoscopique
WO2017007047A1 (fr) Procédé et dispositif de compensation de la non-uniformité de la profondeur spatiale en utilisant une comparaison avec gigue
WO2016006901A1 (fr) Procédé et appareil d'extraction d'informations de profondeur d'une image
WO2022177069A1 (fr) Procédé d'étiquetage et dispositif informatique associé
EP4272458A1 (fr) Procédé et dispositif électronique pour capturer un média à l'aide d'une caméra sous écran
WO2021107592A1 (fr) Système et procédé de retouche d'image précise pour éliminer un contenu non souhaité d'images numériques
WO2022014831A1 (fr) Procédé et dispositif de détection d'objet
WO2023224261A1 (fr) Procédé de dépersonnalisation d'une région liée à la confidentialité dans une image et dispositif de dépersonnalisation l'utilisant
WO2024111931A1 (fr) Correction de tache d'airy pour déflouter une image
WO2020171257A1 (fr) Procédé de traitement d'image et dispositif correspondant
WO2019147028A1 (fr) Appareil de traitement d'image, procédé de traitement d'image et support d'enregistrement lisible par ordinateur

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: 15870112

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: 15870112

Country of ref document: EP

Kind code of ref document: A1