NL2010027C2 - Method for determining clothing sizes of persons from digital pictures. - Google Patents

Method for determining clothing sizes of persons from digital pictures. Download PDF

Info

Publication number
NL2010027C2
NL2010027C2 NL2010027A NL2010027A NL2010027C2 NL 2010027 C2 NL2010027 C2 NL 2010027C2 NL 2010027 A NL2010027 A NL 2010027A NL 2010027 A NL2010027 A NL 2010027A NL 2010027 C2 NL2010027 C2 NL 2010027C2
Authority
NL
Netherlands
Prior art keywords
person
pixels
determining
contrasting
distance
Prior art date
Application number
NL2010027A
Other languages
Dutch (nl)
Inventor
Remco Geert Jan Janssen
Original Assignee
Suit Supply B V
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 Suit Supply B V filed Critical Suit Supply B V
Priority to NL2010027A priority Critical patent/NL2010027C2/en
Priority to PCT/NL2013/050921 priority patent/WO2014098591A1/en
Application granted granted Critical
Publication of NL2010027C2 publication Critical patent/NL2010027C2/en

Links

Classifications

    • A—HUMAN NECESSITIES
    • A41—WEARING APPAREL
    • A41H—APPLIANCES OR METHODS FOR MAKING CLOTHES, e.g. FOR DRESS-MAKING OR FOR TAILORING, NOT OTHERWISE PROVIDED FOR
    • A41H1/00—Measuring aids or methods
    • A41H1/02—Devices for taking measurements on the human body

Landscapes

  • Life Sciences & Earth Sciences (AREA)
  • Biophysics (AREA)
  • Engineering & Computer Science (AREA)
  • Textile Engineering (AREA)
  • Image Analysis (AREA)

Abstract

Method for determining clothing sizes of persons from digital pictures, comprising the following steps in a suitable order: a) taking a frontal digital picture of at least the upper body part of said person while the arms of said person are spread in oblique lateral downward directions, said picture having an upper edge closest to the head of the person and a lower edge farthest from the head of the person; b) defining a horizontal line in said picture which is statistically almost certainly located well below the persons armpits; c) determining on both sides of the vertical centre line of the picture the distances, for instance in pixels, between the pixel of a horizontal line and the first contrasting pixel above said line which has a predetermined significant different colour and/or brightness than the colour and/or brightness of the pixels below it; d) determining on each side of the vertical centre line an uppermost contrasting pixel having a local maximum determined distance from said horizontal line; e) determining the distance, for instance in pixels, between said two uppermost contrasting pixels; f) using said distance between said two uppermost contrasting pixels for calculating a clothing size of said person.

Description

Method for determining clothing sizes of persons from digital pictures
The invention relates to a method for determining clothing 5 sizes of persons from digital pictures. Several such methods already exist, which are for instance used when ordering clothes on the internet. Typically the person ordering the clothes has at least one picture taken (usually two, one frontal and one lateral picture) in his own environment such 10 as home, and sends these pictures together with a reference data such as his length and his order details to the internet shop selling the clothes. However, it appears to be difficult to provide a computer implemented method which automatically determines the correct clothing size of a 15 person form these digital pictures. In particular it appears to be difficult to determine an accurate chest circumference.
The invention aims at providing a method which is better 20 able to determine an accurate chest circumference.
According to the invention, the method comprises the following steps in a suitable order: a) taking a frontal digital picture of at least the 25 upper body part of said person while the arms of said person are spread in oblique lateral downward directions, said picture having an upper edge closest to the head of the person and a lower edge farthest from the head of the person; 30 b) defining a horizontal line in said picture which is statistically almost certainly located well below the persons armpits; 2 c) determining on both sides of the vertical centre line of the picture the distances, for instance in pixels, between the pixel of a horizontal line and the first contrasting pixel above said line which has a predetermined 5 significant different colour and/or brightness than the colour and/or brightness of the pixels below it; d) determining on each side of the vertical centre line an uppermost contrasting pixel having a local maximum determined distance from said horizontal line; 10 e) determining the distance, for instance in pixels, between said two uppermost contrasting pixels; f) using said distance between said two uppermost contrasting pixels for calculating a clothing size of said person.
15
Preferably the method further comprises between step a) and step c), the step of digitally processing said picture such that a contour line of said person is determined. Such processes are well known as such. Preferably the method 20 further comprises the step of filling the interior of said contour line with a monochrome colour. Preferably the method further comprises the step of filling the exterior of said contour line with a different monochrome colour. In this manner, the most accurate high contrast picture can be 25 obtained for use in steps d and e.
Preferably the method further comprises the step of determining the length, for instance in pixels, of said person in said picture, and using the real length, for 30 instance in mm, of the person to convert the distance, for instance in pixels, between said armpits in said picture into an actual distance, for instance in mm, between the armpits of said person. Preferably said method further 3 comprises the step of detecting the face of said person, and using said face recognition data to determine the length of the person. By using face recognition the height of a person's hair dress can be ignored. Said method further 5 preferably comprises the step of using said face recognition data to determine the interior side of said contour.
Said method further preferably comprises the step of taking a lateral digital picture of at least the upper body part of 10 said person and determining the chest depth, for instance in pixels, and using the chest depth in combination with the distance, for instance in pixels, determined in step e for calculating a clothing size of said person in step f.
15 Preferably in step a mobile computer programme is used, which is arranged to be loaded on a mobile computer device comprising a camera and a display for displaying the camera images while taking pictures, such as a smart phone, and which is arranged to display a contour on the display in 20 which the person is expected to be positioned while taking said frontal picture. The contour indicates not only where the person should stand, but also should indicate how the arms must be spread. In this manner the accuracy of the process is enhanced.
25
The invention also relates to a computer programme for determining clothing sizes of persons from digital pictures, arranged to carry out the following steps in a suitable order : 30 a) importing a frontal digital picture of at least the upper body part of a person while the arms of said person are spread in oblique lateral downward directions, said picture having an upper edge closest to the head of the 4 person and a lower edge farthest from the head of the person; b) defining a horizontal line in said picture which is statistically almost certainly located well below the 5 persons armpits; c) determining on both sides of the vertical centre line of the picture the distances, for instance in pixels, between the pixel of a horizontal line and the first contrasting pixel above said line which has a predetermined 10 significant different colour and/or brightness than the colour and/or brightness of the pixels below it; d) determining on each side of the vertical centre line an uppermost contrasting pixel having a local maximum determined distance from said horizontal line; 15 e) determining the distance, for instance in pixels, between said two uppermost contrasting pixels; f) using said distance between said two uppermost contrasting pixels for calculating a clothing size of said person.
20
The invention will now be elucidated by means of a preferred embodiment, as shown in the drawings, wherein:
Figure 1 is a frontal digital picture of a person; and 25
Figures 2-10 show various stages of the digital picture while being processed.
Figure 1 shows a frontal digital picture of a man in 30 underwear. This picture may be taken at home with a smart phone having a camera and a display, and using software which is arranged to provide instructions as to how to take the required pictures for ordering clothes online. The 5 instructions include instructions to take one frontal picture and one lateral picture of the person for whom the clothes are ordered. The instructions also comprise contours on the display of the smart phone for each picture in which 5 the person is expected to be positioned while the picture is being taken. Such a contour for the frontal picture is shown in figure 2. The person is instructed to hold his arms as shown, i.e. spread laterally in oblique downward directions. The process of taking the lateral picture is not shown, but 10 very similar. The person is instructed to hold his arms straight down next to his upper body when taking the lateral picture .
After the pictures have been taken, the person can upload 15 the pictures to the website of an online shop, together with his order, for instance a shirt or a suit. While filling in the order, the person is also asked to fill in his real length.
20 The received digital pictures and data are then digitally and automatically processed as follows. As shown in figure 3, the digital picture of figure 1 is digitally processed to provide a black on white contour image by using contrast comparison techniques (the image may of course also be white 25 on black, or any other combination of colours). Such processes are well known as such, and need no further explanation here. As shown in figure 4 the pre-defined contour as shown in figure 2 is used again to filter the image, such that all black pixels outside the pre-defined 30 contour are deleted from the image. In the step as shown in figure 5, the black pixels in the image are made bigger by rendering the neighbouring pixels also black. This ensures that a closed contour is obtained. In figure 6 the thus 6 obtained contour is made thinner again, while ensuring that the contour remains closed. In figure 7, the inner side of the contour is filled with black pixels. In order to determine the inner side of the contour, a face recognition 5 algorithm may be used on the original picture of figure 1. But the inner side of the contour may also be guessed by using a predetermined pixel in the image as the starting point for filling the contour.
10 In figure 8 a narrowest vertical rectangle is drawn, which touches the outer black pixels on all four sides. In figure 9 a horizontal reference line is drawn at a predetermined height in the rectangle, which statistically is at an average person's hip height.
15
Figure 10 shows the process to determine the location of the person's armpits, which is carried out on both sides of the vertical rectangle's vertical centre line. The computer programme is arranged to determine for each black pixel on 20 the reference line, the number of white pixels between the pixel on reference line and the nearest black pixel in vertical direction (as shown by the arrows). The computer programme is further arranged to determine for each side the maximum number of white pixels thus determined, and the 25 uppermost black pixel related therewith. The two pixels determined in this manner represent the locations of the person's armpits.
The computer is further arranged to determine the distance 30 in pixels between the two "arm pit" pixels, and to use a chest depth determined from the second, lateral digital picture, and the person's real length as provided by the user and the person's length in pixels determined by the 7 height of the vertical rectangle of figure 8, to calculate a "real world" chest size, and therewith a clothing size.
In order to further improve the calculation, the height of 5 the person derived from the height of the vertical rectangle of figure 8 is corrected by using face recognition techniques in order to estimate the location of the upper side of the head, instead of the upper side of the hair dress. Also a correction can be made, for instance by a 10 predetermined factor, for the extra height added to the vertical rectangle by the fact that the outer ends of the feet of the person appear in the picture below the lower side of the heel.
15 The invention has thus been described by means of preferred embodiments. It is to be understood, however, that this disclosure is merely illustrative. Various details of the structure and function were presented, but changes made therein, to the full extent extended by the general meaning 20 of the terms in which the appended claims are expressed, are understood to be within the principle of the present invention. The description and drawings shall be used to interpret the claims. The claims should not be interpreted as meaning that the extent of the protection sought is to be 25 understood as that defined by the strict, literal meaning of the wording used in the claims, the description and drawings being employed only for the purpose of resolving an ambiguity found in the claims. For the purpose of determining the extent of protection sought by the claims, 30 due account shall be taken of any element which is equivalent to an element specified therein.

Claims (10)

1. Werkwijze voor het bepalen kledingmaten personen uit digitale foto's omvattende de volgende stappen in een 5 geschikte volgorde: a) het nemen van een frontale digitale foto van ten minste het bovenste lichaamsdeel van genoemde persoon terwijl de armen van de persoon zich uitstrekken in schuin zijwaartse en neerwaartse richtingen, waarbij de foto een bovenrand 10 heeft die het dichtst bij het hoofd van de persoon ligt en een onderrand die het verst van het hoofd van de persoon ligt; b) het definiëren van een horizontale lijn in de foto die statistisch vrijwel zeker ruim onder de oksels van de 15 persoon is gelegen; c) het aan weerszijden van de verticale hartlijn van het beeld bepalen van de afstanden, bijvoorbeeld in pixels tussen de pixel van de horizontale lijn en de eerste contrasterende pixel boven de lijn die een vooraf bepaalde 20 significant andere kleur en/of helderheid heeft dan de kleur en/of helderheid van de pixels eronder; d) het bepalen aan weerszijden van de verticale hartlijn van een bovenste contrasterende pixel die een lokale maximum bepaalde afstand van de horizontale lijn heeft; 25 e) het bepalen van de afstand, bijvoorbeeld in pixels tussen de genoemde twee bovenste contrasterende pixels; f) het gebruiken van de afstand tussen de twee bovenste contrasterende pixels voor de berekening van een kledingmaat van de persoon. 30Method for determining clothing sizes of persons from digital photos comprising the following steps in a suitable order: a) taking a frontal digital photo of at least the upper body part of said person while the person's arms extend obliquely sideways and downward directions, wherein the photograph has an upper edge 10 closest to the person's head and a lower edge furthest away from the person's head; b) defining a horizontal line in the photograph that is almost certainly well below the armpits of the person; c) determining the distances on either side of the vertical center line of the image, for example in pixels between the pixel of the horizontal line and the first contrasting pixel above the line that has a predetermined significantly different color and / or brightness than the color and / or brightness of the pixels below; d) determining on either side of the vertical center line of an upper contrasting pixel that has a local maximum determined distance from the horizontal line; E) determining the distance, for example in pixels, between said two upper contrasting pixels; f) using the distance between the two upper contrasting pixels for the calculation of a person's clothing size. 30 2. Werkwijze volgens conclusie 1, verder omvattende de stap van het, tussen stap a) en stap c), digitaal verwerken van de afbeelding zodat een contourlijn van de persoon wordt bepaald.The method of claim 1, further comprising the step of digitally processing the image between step a) and step c) so that a contour line of the person is determined. 3. Werkwijze volgens conclusie 2, verder omvattende de stap 5 van het vullen het inwendige van de contourlijn met een monochrome kleur.The method of claim 2, further comprising the step of filling the interior of the contour line with a monochrome color. 4. Werkwijze volgens conclusie 2 of 3, verder omvattende de stap van het vullen van het uitwendige van de contourlijn 10 met een andere monochrome kleur.The method of claim 2 or 3, further comprising the step of filling the exterior of the contour line 10 with a different monochrome color. 5. Werkwijze volgens een van de voorgaande conclusies, verder omvattende de stap van het bepalen van de lengte, bijvoorbeeld in pixels, van de persoon in de foto, en het 15 door middel van de werkelijke lengte, bijvoorbeeld in mm, van de persoon converteren van de afstand bijvoorbeeld in pixels, tussen de oksels in de foto in een werkelijke afstand, bijvoorbeeld in mm, tussen de oksels van deze persoon. 205. Method as claimed in any of the foregoing claims, further comprising the step of determining the length of the person in the photo, for example in pixels, and converting the person by means of the actual length, for example in mm, of the person of the distance, for example in pixels, between the armpits in the photo in an actual distance, for example in mm, between the armpits of this person. 20 6. Werkwijze volgens een van de voorgaande conclusies, verder omvattende de stap van detecteren van het gezicht van de persoon, en onder gebruikmaking van de gezichtsherkenningsgegevens het bepalen van de lengte van de 25 persoon.6. Method as claimed in any of the foregoing claims, further comprising the step of detecting the face of the person, and using the face recognition data to determine the length of the person. 7. Werkwijze volgens een van de voorgaande conclusies, verder omvattende de stap van detecteren van het gezicht van de persoon, en onder gebruikmaking van de 30 gezichtsherkenningsgegevens het bepalen de binnenzijde van de contour.7. Method as claimed in any of the foregoing claims, further comprising the step of detecting the face of the person, and determining the inside of the contour using the face recognition data. 8. Werkwijze volgens een van de voorgaande conclusies, verder omvattende de stap van het nemen van een laterale digitale foto van ten minste het bovenste lichaamsdeel van de persoon en het bepalen van de borstdiepte, bijvoorbeeld 5 in pixels, en onder gebruikmaking van de borstdiepte in combinatie met de afstand, bijvoorbeeld in pixels, bepaald in stap e, het berekenen van een kledingmaat van deze persoon in stap f.Method according to any of the preceding claims, further comprising the step of taking a lateral digital photo of at least the upper body part of the person and determining the breast depth, for example 5 in pixels, and using the breast depth in combination with the distance, for example in pixels, determined in step e, calculating a clothing size of this person in step f. 9. Werkwijze volgens een van de voorgaande conclusies, waarbij in stap a een programma voor een mobiele computer wordt gebruikt, dat is ingericht om te worden geladen op een mobiele computer omvattende een camera en een scherm voor het weergeven van de camerabeelden tijdens het maken van 15 foto's, zoals een smartphone, en dat is ingericht om een contour weer te geven op het scherm waarin de persoon wordt geacht plaats te nemen terwijl de frontale foto wordt genomen.A method according to any one of the preceding claims, wherein in step a a program for a mobile computer is used, which is arranged to be loaded on a mobile computer comprising a camera and a screen for displaying the camera images during the making of 15 photos, such as a smartphone, which is arranged to display a contour on the screen in which the person is supposed to take a seat while the frontal photo is taken. 10. Computerprogramma voor het bepalen van kledingmaten van personen uit digitale foto's, dat is ingericht voor het uitvoeren van de volgende stappen in een geschikte volgorde: a) het importeren van een frontale digitale foto van ten minste het bovenste lichaamsdeel van een persoon terwijl de 25 armen van de persoon uitstrekken zich in schuin zijwaartse en neerwaartse richtingen, waarbij de foto een bovenrand heeft die het dichtst bij het hoofd van de persoon ligt en een onderrand die het verst van het hoofd van de persoon ligt; 30 b) het definiëren van een horizontale lijn in de foto die statistisch vrijwel zeker ruim onder de oksels van de persoon is gelegen; c) het aan weerszijden van de verticale hartlijn van het beeld bepalen van de afstanden, bijvoorbeeld in pixels tussen de pixel van een horizontale lijn en de eerste contrasterende pixel boven de lijn die een vooraf bepaalde 5 significant andere kleur en/of helderheid heeft dan de kleur en/of helderheid van de pixels eronder; d) het aan weerszijden van de verticale hartlijn bepalen van een bovenste contrasterende pixel die een lokale maximum bepaalde afstand van de horizontale lijn heeft; 10 e) het bepalen van de afstand, bijvoorbeeld in pixels, tussen de genoemde twee bovenste contrasterende pixels; f) het gebruiken van de afstand tussen de twee bovenste contrasterende pixels voor de berekening van een kledingmaat van de persoon.10. Computer program for determining clothing sizes of persons from digital photos, which is arranged for performing the following steps in a suitable order: a) importing a frontal digital photo of at least the upper body part of a person while the the person's arms extend in oblique lateral and downward directions, the photograph having an upper edge closest to the person's head and a lower edge furthest away from the person's head; B) defining a horizontal line in the photo that is almost certainly well below the armpits of the person; c) determining the distances on either side of the vertical center line of the image, for example in pixels between the pixels of a horizontal line and the first contrasting pixel above the line which has a predetermined significantly different color and / or brightness than the color and / or brightness of the pixels below; d) determining on either side of the vertical center line an upper contrasting pixel that has a local maximum determined distance from the horizontal line; E) determining the distance, for example in pixels, between said two upper contrasting pixels; f) using the distance between the two upper contrasting pixels for the calculation of a person's clothing size.
NL2010027A 2012-12-20 2012-12-20 Method for determining clothing sizes of persons from digital pictures. NL2010027C2 (en)

Priority Applications (2)

Application Number Priority Date Filing Date Title
NL2010027A NL2010027C2 (en) 2012-12-20 2012-12-20 Method for determining clothing sizes of persons from digital pictures.
PCT/NL2013/050921 WO2014098591A1 (en) 2012-12-20 2013-12-19 Method for determining clothing sizes of persons from digital pictures

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
NL2010027A NL2010027C2 (en) 2012-12-20 2012-12-20 Method for determining clothing sizes of persons from digital pictures.
NL2010027 2017-12-06

Publications (1)

Publication Number Publication Date
NL2010027C2 true NL2010027C2 (en) 2014-06-23

Family

ID=48142872

Family Applications (1)

Application Number Title Priority Date Filing Date
NL2010027A NL2010027C2 (en) 2012-12-20 2012-12-20 Method for determining clothing sizes of persons from digital pictures.

Country Status (2)

Country Link
NL (1) NL2010027C2 (en)
WO (1) WO2014098591A1 (en)

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US5515168A (en) * 1991-06-03 1996-05-07 Telmat Informatique, E.U.R.L. Identification and classification device and process practiced with said device
US6490534B1 (en) * 2000-04-25 2002-12-03 Henry Pfister Camera measurement system
WO2008145952A1 (en) * 2007-05-29 2008-12-04 Sony Computer Entertainment Europe Limited Apparatus and method of body characterisation
WO2011142655A2 (en) * 2010-05-10 2011-11-17 Suitsupply B.V. Method for remotely determining clothes dimensions

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US5515168A (en) * 1991-06-03 1996-05-07 Telmat Informatique, E.U.R.L. Identification and classification device and process practiced with said device
US6490534B1 (en) * 2000-04-25 2002-12-03 Henry Pfister Camera measurement system
WO2008145952A1 (en) * 2007-05-29 2008-12-04 Sony Computer Entertainment Europe Limited Apparatus and method of body characterisation
WO2011142655A2 (en) * 2010-05-10 2011-11-17 Suitsupply B.V. Method for remotely determining clothes dimensions

Also Published As

Publication number Publication date
WO2014098591A1 (en) 2014-06-26

Similar Documents

Publication Publication Date Title
US10964078B2 (en) System, device, and method of virtual dressing utilizing image processing, machine learning, and computer vision
EP3038053B1 (en) Method and system for generating garment model data
US10620005B2 (en) Building height calculation method, device, and storage medium
JP6554169B2 (en) Object recognition device and object recognition system
WO2016000304A1 (en) Virtual fitting method and virtual fitting system
WO2020239251A1 (en) Methods and systems for determining body measurements and providing clothing size recommendations
US20200074667A1 (en) Clothing Size Determination Systems and Methods of Use
US11048963B2 (en) Product image generation system
CN107767325A (en) Video processing method and device
WO2016051416A1 (en) System and method of 3d modeling and virtual fitting of 3d objects
CN105930762A (en) Eyeball tracking method and device
US12086871B2 (en) Digital imaging systems and methods for detecting user-specific apparel fit
JP7071086B2 (en) Image processing equipment, image processing methods and computer programs
CN115039150B (en) Judgment method, judgment device, and computer program product
KR102328072B1 (en) Server for providing virtual fitting service and method using the same
US20160071321A1 (en) Image processing device, image processing system and storage medium
JP2014071501A (en) Image processor, image processing method and program
JP5868426B2 (en) How to estimate the orientation of a stationary person
US20180096490A1 (en) Method for determining anthropometric measurements of person
NL2010027C2 (en) Method for determining clothing sizes of persons from digital pictures.
JP7529044B2 (en) Image processing device, image processing method, and program
JP6404794B2 (en) Image information acquisition method, image evaluation method, image information acquisition device, image evaluation device, and image processing program
JP4272538B2 (en) Moving object detection apparatus, moving object detection method, and moving object detection program
JP7024876B2 (en) Detection device, processing device, detection method, and processing program
JP4775599B2 (en) Eye position detection method