Systems and methods for optical recognition of tire specification

• US 9,798,946 B2
• Filed: 11/13/2015
• Issued: 10/24/2017
• Est. Priority Date: 01/29/2015
• Status: Active Grant

• Current Assignee: DAIN Company Limited
• Original Assignee: DAIN Company Limited
• Inventors: Lee, Young Gi, Lee, Kun Woo, Baek, Seung Yeob, Kam, Dong Uk, Kim, Dae Wook, Yoo, Soo Gon
• Primary Examiner(s): YEH, EUENG NAN
• Application Number:
  US14/941,313
• Publication Number:
  US 20160224863A1
• Time in Patent Office: 711 Days
• CPC Class Codes: G06V 30/10 Character recognition
• CPC Class Codes: G06V 30/1456 based on user interactions
• CPC Class Codes: G06V 30/1475 Inclination or skew detection or correction of characters or of image to be recognised
• CPC Class Codes: G06V 30/1607 Correcting image deformation, e.g. trapezoidal deformation caused by perspective

Abstract

Provided are a method and an apparatus for recognizing a tire by using an image of a tire captured by using a terminal. The apparatus recognizes a tire by converting an image of a round tire included in the image into a linear image, extracting an area of a character to be recognized from the linear image, and determining a character for learning, which is most similar to the extracted area of the character to be recognized from among a pre-constructed group of characters for learning, as a character in the area of the character to be recognized.

Claims

1. A tire recognition method comprising:

receiving an image of a round tire using a camera of a mobile device;

converting, by the mobile device, the image of the round tire included in the image into a linear image of the round tire by;

extracting, by the mobile device, the image of the round tire from the received image; and

obtaining, by the mobile device, the linear image of the round tire by rearranging linear images in a same direction, wherein each of the linear images corresponds to at least one line that is perpendicular to a circumference of a circle or an oval corresponding to a wheel in the tire;

extracting, by the mobile device, an area of a character to be recognized from the linear image;

determining, by the mobile device, a character for learning, which is most similar to the extracted area of the character to be recognized from among a pre-constructed group of characters for learning, as a character in the area of the character to be recognized; and

retrieving, by the mobile device, information associated with the round tire based on the character for learning.

2. The tire recognition method of claim 1, wherein the converting of the image of the round tire into the linear image further comprises:

removing a wheel from the image of the round tire prior to obtaining the linear image.

3. The tire recognition method of claim 1, wherein extracting the image of the round tire comprises:

extracting a certain area, which includes the round tire, from the received image of the round tire based on at least one guideline provided when the received image of the round tire was captured prior to obtaining the linear image.

4. The tire recognition method of claim 3, wherein the extracting of the area further comprises extracting a certain area located between two guidelines, which were provided when the received image was captured and have a diameter different from each other, from the received image.

5. The tire recognition method of claim 1, wherein:

the image of the round tire is extracted based on a guideline provided when the received image was captured;

converting the image further comprises;

obtaining a boundary between the round tire and a wheel based on intensity in the linear image;

extracting an additional image of the round tire from the received image based on the boundary; and

obtaining, by the mobile device, a new linear image of the round tire by rearranging new linear images from the extracted additional image in a same direction, wherein each of the new linear images corresponds to at least one line that is perpendicular to a circumference of a circle or an oval corresponding to a wheel in the tire.

6. The tire recognition method of claim 5, wherein the obtaining of the boundary between the wheel and the tire comprises:

providing a weight value to each line in the linear image according to a direction of a boundary of each respective line;

obtaining additional boundaries between each area which is divided based on intensity in the linear image; and

obtaining a shortest path connecting both ends of a path image, which consists of the additional boundaries between each area, to each other, by applying the weight value for each respective line.

7. The tire recognition method of claim 6, wherein the secondarily converting of the image comprises:

converting the shortest path into a round image;

correcting a shortest path in the round image to have a form of a circle or an oval;

re-extracting a tire in the received image based on the circle or the oval obtained by the correcting; and

obtaining, by the mobile device, a new linear image of the round tire by rearranging new linear images from the re-extracted tire in the received image in a same direction, wherein each of the new linear images corresponds to at least one line that is perpendicular to a circumference of a circle or an oval corresponding to a wheel in the tire.

8. The tire recognition method of claim 6, wherein the extracting of the area of a character to be recognized comprises extracting the character area from the linear image by using at least one feature from the group consisting of texture, a pattern, and an edge.

9. The tire recognition method of claim 1, wherein the extracting of the area of a character to be recognized comprises:

applying an entropy filter to the linear image;

obtaining a binary image consisting of black and white areas by using an adaptive threshold in an image obtained by applying the entropy filter to the linear image;

removing a noise area, whose size is smaller than a predetermined size, from the binary image; and

extracting an area of a character to be recognized from an image obtained by removing noise from the binary image.

10. The tire recognition method of claim 1, wherein the determining of the character for learning comprises:

obtaining feature information by applying at least one filter to the area of a character to be recognized; and

determining a character for learning as a character included in the area of a character to be recognized, the character for learning having feature information that is most similar to feature information about the area of a character to be recognized, based on comparison between pre-constructed feature information stored within a database and the feature information about the area of a character to be recognized.

11. The tire recognition method of claim 10, wherein the at least one filter is a filter that may extract a feature, such as a Haar-like filter or a gabor filter.

12. A non-transitory computer-readable recording storage medium having recorded thereon a computer program which, when executed by a computer, performs the method of claim 1.