DE102018108284A1 - A method of measuring the size of an object by processing its image - Google Patents

Abstract

A method is proposed for measuring the distance of two points of an object by processing its image, wherein the image was taken with a mobile phone. For example, the article (eg, a human foot) stands on a rectangular paper (1) with a mark (11) in each corner. The picture must show the foot and at least three marks (11). A person stands next to the paper (1), which is attached to a wall, whereby with this method the paper (1) can be pressed with one hand. The measured distance can be used to obtain a correct size of the foot.

Description

Technical area

The invention relates to a method of measuring the size of an object (e.g., a human foot) by processing its image, the user standing on a rectangular paper with a mark in each corner. In the drawing, three of these marks are shown.

State of the art

Usually, an online shopper looks for an e-catalog when searching for a product. For shoes, for example, the buyer measures the foot and selects the appropriate size in the e-catalog before he or she places the order. However, it often happens that the buyer sacrifices a lot of time and effort for the measurement. In addition, the size varies with age. Saving the history of sizes is difficult. Furthermore, the ordered goods, such as shoes, may not be the right ones due to a visual error. This is annoying for both the seller and the buyer. This situation needs further improvement.

Object of the invention

An object of the invention is to provide a method of measuring the size of an object, comprising the steps of (a) taking an image of an object with a cell phone, the object resting on a rectangular paper with four markings in each of the four corners, to get a picture of the object; Reducing the size of the image and visually balancing the image; Comparing the map with a predetermined value to filter out the noise; further comparing the map with an adaptive threshold to filter out the noise; Application of intelligent edge detection to illustrate an imaging edge; Calculation of the number of outlines of the mark; Calculation of the number of marks in the figure; Marking the three marks shown in the figure as first, second and third marks with one coordinate; Obtaining a fourth mark using a projection; Obtaining a location of the object in three-dimensional space; Performing a perspective transformation on the object to create a distance between any two points of the object in the image; Subjugation of the mapping of a transformation of the Lab color space; Adjustment of image size; Converting the two-dimensional properties of the image into its one-dimensional property; Perform clustering with k means (k = 3) to obtain cluster centers; Applying the nearest neighbor classifier at the cluster centers to classify the new data into bundles and calculate the number of bundles using the Asian color index (L: 90, a: 12, b: 12); Binarizing the skin color of the image to convert a pixel image to a binary image; Measuring the length of the object or its width; Performing the perspective transformation on the object in the image; and getting a proper size of the object.

The above objects, features and advantages of the invention will be more fully understood from the following detailed description with reference to the accompanying drawings.

list of figures

• 1 shows a schematic illustration of a foot standing on a rectangular paper with a mark in each corner according to a first embodiment of the present invention;
• 2 FIG. 10 illustrates a flowchart of a method of selecting a shoe by manipulating the image of the foot according to a first embodiment of the present invention; FIG.
• 3 shows a schematic illustration of the six outlines of the mark;
• 4 shows a schematic illustration of the method for finding the fourth mark on the paper;
• 5 shows the length of the foot in the picture;
• 6 shows the width of the foot in the picture;
• 7 shows a schematic illustration of the foot, which stands on a rectangular paper with three marks in each one of the three corners and a fourth mark on a rod, this fourth mark is on a rod, which rests on the paper and next to the toes the foot is located according to a second embodiment of the present invention;
• 8th FIG. 12 illustrates a flowchart of a method of selecting a shoe by manipulating the image of the foot according to the second embodiment of the present invention; FIG. and
• 9 shows a schematic illustration of a person standing in front of a wall with a this adhered rectangular paper is standing, wherein the paper has a mark in each corner according to a third embodiment of the present invention.

Ways to carry out the invention

The 1 shows an object (eg the foot) standing on a rectangular piece of paper 1 with a marker 11 in each corner according to a first embodiment of the present invention.

• (A1) Aufnahme eines Bildes des in der 1 gezeigten Fußes mit einem Handy, um eine Abbildung des Fußes zu erhalten, wobei der Fuß auf derselben Ebene wie die Markierungen 11 stehen muss;
• (A2) Verkleinerung der Größe der Abbildung und visuelles Ausgleichen des Bildes;
• (A3) Vergleich der Abbildung mit einem vorbestimmten Wert, um das Rauschen herauszufiltern, sowie weiterer Vergleich der Abbildung mit einem anpassungsfähigen Schwellenwert, um das Rauschen herauszufiltern und eine deutlichere Abbildung zu erhalten;
• (A4) Anwendung einer intelligenten Kantendetektion zur Verdeutlichung eines Abbildungsrands;
• (A5) Berechnung der Anzahl der Umrisse der Markierung 11, wobei die Anzahl der Umrisse wegen den sechs Pfeilen in der 3 sechs beträgt;
• (A6) Berechnung der Anzahl der Markierungen 11 in der Abbildung;
• (A7) ein Flußdiagramm für den Übergang zum nächsten Schritt, falls die Anzahl der Markierungen 11 drei beträgt, da sonst das Flußdiagramm zu Schritt (A1) zurückkehrt;
• (A8) wie in 4 gezeigt, wird die Markierung 11 oben links als eine erste Markierung mit einer ersten Koordinate, die Markierung 11 oben links als eine zweite Markierung 11 mit einer zweiten Koordinate und die Markierung 11 oben rechts wird als eine dritte Markierung mit einer dritten Koordinate definiert, so dass mit Hilfe einer Hochrechnung eine vierte Markierung mit einer vierten Koordinate in der unteren rechten Ecke erhalten werden kann;
• (A9) Erhalten des Standorts der Abbildung im dreidimensionalen Raum und Ausführen einer perspektivischen Transformation, um einen Abstand zwischen zwei beliebigen Punkten in der Abbildung zu schaffen;
• (B1) Unterwerfung der Abbildung einer Transformation des Lab-Farbraums, da der RGB-Farbraum die Hautfarbe nur sehr begrenzt identifiziert, wobei die Hautfarbe mit dem Lab-Farbraum verstärkt werden kann;
• (B2) Anpassung der Bildgröße und Umwandlung der zweidimensionalen Eigenschaften der Abbildung in deren eindimensionale Eigenschaft;
• (B3) Ausführen der Bündelung mit k-Mitteln (k=3), um Clusterzentren zu erhalten, wobei die Farbe als eine erste Bündelung der Hautfarbe, eine zweite Bündelung der weißen Farbe und eine dritte Bündelung des Hintergrunds klassifiziert werden;
• (B4) Erhalten einer Anzahl von Bündelungen und Clusterzentren, insbesondere des erstnächsten benachbarten Klassierers der Clusterzentren, der durch k-Mittel erhalten wurde und zum Klassifizieren der neuen Daten in den vorhandenen Bündelungen verwendet wird, wobei das Berechnen der Anzahl der Bündelungen und Clusterzentren mit Hilfe der asiatische Farbindex (L:90, a:12, b:12) erfolgt;
• (B5) Binarisierung der Hautfarbe der Abbildung zum Umwandeln eines Pixelbilds in ein Binärbild, insbesondere wird ein Abstand zwischen jedem Pixel der Abbildung und der nächsten Bündelung als 255 und ein Abstand zwischen derselben Bündelung und dem distalen Clusterzentrum als 0 definiert;
• (B6) Messung der Fußlänge (5) bzw. der Fußbreite (6);
• (C1) Ausführung einer perspektivischen Transformation des Fußes in der Abbildung;
• (C2) Erhalten der richtigen Größe des Fußes;
• (D1) Eingabe der Größe des Fußes in eine Schuhdatenbank, in dem zahlreiche verschiedene Größen und Schuharten gespeichert sind; und
• (D2) Suche der Schuhdatenbank, um einen geeigneten nummerierten Schuh, in den der Fuß passt, zu finden.

The 2 to 6 show in connection with 1 a method of selecting a shoe by processing the image according to a first embodiment of the present invention. This procedure includes the following steps:

• ( A1 ) Taking a picture of the in the 1 foot with a mobile phone to get an image of the foot, with the foot on the same level as the markers 11 must stand;
• ( A2 ) Reducing the size of the image and visually balancing the image;
• ( A3 ) Compare the image with a predetermined value to filter out the noise, and further compare the image with an adaptive threshold to filter out the noise and obtain a clearer image;
• ( A4 ) Application of intelligent edge detection to illustrate an imaging edge;
• ( A5 ) Calculation of the number of outlines of the marking 11 , the number of outlines because of the six arrows in the 3 is six;
• ( A6 ) Calculation of the number of marks 11 in the picture;
• ( A7 ) a flow chart for the transition to the next step, if the number of marks 11 is three, otherwise the flowchart for step ( A1 ) returns;
• ( A8 ) as in 4 shown, the mark becomes 11 top left as a first marker with a first coordinate, the marker 11 top left as a second mark 11 with a second coordinate and the marker 11 upper right is defined as a third marker with a third coordinate, so that by means of a projection a fourth marker with a fourth coordinate in the lower right corner can be obtained;
• ( A9 ) Obtaining the location of the image in three-dimensional space and performing a perspective transformation to create a space between any two points in the image;
• ( B1 ) Subjugation of the mapping of a transformation of the Lab color space, since the RGB color space identifies skin color only to a very limited extent, whereby the skin color can be intensified with the Lab color space;
• ( B2 ) Adaptation of the image size and conversion of the two-dimensional properties of the image into its one-dimensional property;
• ( B3 ) Performing clustering with k means (k = 3) to obtain cluster centers, classifying the color as a first bundling of the skin color, a second bundling of the white color, and a third bundling of the background;
• ( B4 Obtaining a number of bundles and cluster centers, in particular the cluster cluster first adjacent clusterer, obtained by k-resources and used to classify the new data in the existing bundles, calculating the number of bundles and cluster centers using the Asian cluster Color index (L: 90, a: 12, b: 12);
• ( B5 Binarization of the skin color of the image to convert a pixel image into a binary image, in particular a distance between each pixel of the image and the next bundle is defined as 255 and a distance between the same bundle and the cluster center distal is 0;
• ( B6 ) Measurement of foot length ( 5 ) or the foot width ( 6 );
• ( C1 ) Performing a perspective transformation of the foot in the figure;
• ( C2 ) Getting the right size of the foot;
• ( D1 ) Enter the size of the foot in a shoe database, which stores many different sizes and types of shoes; and
• ( D2 ) Search the shoe database to find a suitable numbered shoe to fit the foot.

The 7 shows that the foot is on the rectangular paper 1 with three marks 11 in each one of the three corners and with a fourth mark 11 on a pole 12 on the paper 1 and rests next to the toes of the foot according to a second embodiment of the present invention stands.

The 8th FIG. 12 illustrates a flowchart of a method of selecting a shoe by manipulating the image of the foot according to a second embodiment of the present invention. The steps of the method correspond to those described in the 2 are shown.

The 9 Shows a person standing in front of a wall with a glued rectangular paper 1 stands, taking the paper 1 a mark 11 in each corner according to a third embodiment of the present invention.

In spite of the description of the invention in terms of the preferred embodiments, it will be apparent to those skilled in the art that modifications can be made to the invention within the spirit and scope of the following appended claims.

Claims (2)

A method of measuring the size of an article, comprising the steps of: (a) taking an image of an object with a mobile phone, the object lying on a rectangular paper (1) with four markings (11) in each one of the four corners to obtain a respective image of the object; (b) reducing the size of the image and visually equalizing the image; (c) comparing the map with a predetermined value to filter out the noise; (d) further comparing the map with an adaptive threshold to filter out the noise; (e) applying intelligent edge detection to illustrate an imaging edge; (f) calculating the number of outlines of the mark (11), the number of outlines being six; (g) calculating the number of marks (11) in the figure; (h) returning to step (a) if the number of marks (11) is less than three; (i) marking the three marks shown in the figure as first, second and third marks (11) with a coordinate; (j) obtaining a fourth mark using a projection; (k) obtaining a location of the object in three-dimensional space; (I) performing a perspective transformation on the object to provide a distance between any two points of the object in the image; (m) subjecting the mapping to a transformation of the Lab color space; (n) adjusting the image size; (o) converting the two-dimensional properties of the image into its one-dimensional property; (p) performing clustering with k means (k = 3) to obtain cluster centers, classifying the color as a first bundling of the skin color, a second bundling of the white color, and a third bundling of the background; (q) applying the nearest neighbor classifier at the cluster centers to classify the new data into bundles and calculating the number of bundles using the Asian color index (L: 90, a: 12, b: 12); (r) binarizing the skin color of the image to convert a pixel image to a binary image, defining a distance between each pixel of the image and the next bundle as 255 and a distance between the same bundle and the cluster center distal to 0; (s) measuring the length of the object or its width; (t) performing the perspective transformation on the object in the image; and (u) Obtain a correct size of the object. Method according to Claim 1 further comprising a rod (12) having a fifth mark thereon.

Images