WO2019084385A1

WO2019084385A1 - Tire code reader

Info

Publication number: WO2019084385A1
Application number: PCT/US2018/057684
Authority: WO
Inventors: Roger Tracy; Cesare Santanera; David Boyle; Jeffrey HISLOP
Original assignee: Tire Profiles LLC
Current assignee: Tire Profiles LLC
Priority date: 2017-10-26
Filing date: 2018-10-26
Publication date: 2019-05-02
Anticipated expiration: 2020-04-26

Abstract

Disclosed herein is a system and method for reading DOT tire codes. The system includes a camera unit in a palm-sized housing. The housing contains a pair of adjacent imaging devices and plural banks of LED lighting deployed around the camera unit's window to illuminate the tire code from different angles. The camera unit window faces the tire sidewall and over the tire code to obscure external light and obtain a close view of the tire code. Then the banks of lights illuminate the tire code in a pre-determined sequence to enable the imaging devices to capture a series of images of using a neural network to use for character recognition and tire code display.

Description

TITLE

TIRE CODE READER

TECHNOLOGICAL FIELD:

[0001 ] The present disclosure relates generally to devices for recognizing

alphanumeric characters and, in particular, alphanumeric characters on the sidewalls of tires.

BACKGROUND

[0002] Reference is made to US 9,454,707, which is owned by the present

applicant, is intended for use in reading United States Department of

Transportation "DOT" tire identification codes, and is incorporated in its entirety into the present application. Tire Identification codes are embossed or molded into the sidewalls of tires sold in the United States. Other types of tire markings may be used in other countries and tires destined for use in the US may be manufactured with DOT tire codes. These codes identify information relevant to the tire with alphanumeric characters arranged in three sets of four characters. The characters, being black on a black surface, are difficult to read under normal lighting, and even more difficult to read in dim lighting, such as in a garage.

[0003] A device that could reliably read a DOT tire code electronically would be advantageous.

SUMMARY

[0004] Disclosed herein is system and method for reading DOT tire codes

electronically. The term "reading" in this sense means that the individual alphanumeric characters in the sequence they appear on the tire sidewall are detected and replicated electronically. Thus, for example, the letter "A" and the number "9" embossed or molded into the sidewall of a particular tire would be detected in images of the tire sidewall by a programmed computer and

interpreted as the letter A and the number 9 of a tire code, and included with l other alphanumeric characters comprising the tire code from that tire. The system includes a camera unit having one or two cameras in a palm-sized housing. The camera unit captures a series of images using plural lights that sequentially illuminate the tire code from different angles. The sequence of images is made using only artificial light emanating from within the camera and is a fixed, particular sequence. The sequence of images captured by the camera unit are forwarded wirelessly to a server for storage and to a processor for detection and reading of the alphanumeric DOT code of the particular tire. The DOT tire code is displayed for the user.

[0005] A feature of the disclosure is the use of at least two lighting elements, to be used in sequence to light the tire code. Each lighting element used is directed at the tire code from a different angle in order to highlight different features of the topography of the code to improve tire code detection and reading accuracy.

[0006] A feature of the disclosure is that banks of light-emitting diodes (LEDs) may be used for their small size, brightness, low energy consumption and their convenience for directing the light where desired. Moreover, the imaging devices are closer to the tire code when the images are acquired.

[0007] Another feature of the disclosure is that the camera unit contains one or two solid-state cameras (e.g., a charge-coupled device (CCD) or complementary metal-oxide-semiconductor (CMOS) in side-by-side relationship with fields of view that overlap to capture the DOT tire code in a small housing, which can acquire the requisite set of images whether held right side up or upside down. The cameras may have one mega-pixels each for images having good resolution, but the method may need higher resolution to read very small characters.

[0008] A feature of the disclosure is that each camera has its own lens and the housing has a window protecting both cameras.

[0009] A feature of the disclosure is that the camera unit uses a server and

processor away from the camera unit so as to be able to access cloud computing (but any sufficiently powerful computer could be satisfactory), to thereby minimize the size of the camera unit and make more computing power available. [0010] Another feature of the disclosure is that a neural network may be used to extract the DOT codes with greater speed and accuracy. In particular, a convolutional neural network, which is a special kind of multi-layer neural network, may be used. Like many other neural networks, computers are programmed to self-train based on a back-propagation algorithm. Convolutional neural networks, such as LeNet-5, are designed to recognize visual patterns directly from pixel images with minimal programming. Such neural networks can learn to recognize patterns such as hand-written characters with extreme variability, and are robust against distortions and simple geometric

transformations.

[001 1 ] These and other features and their advantages will be apparent to those skilled in the art of tire inventory management from a careful readying of the detailed description accompanied by the following drawings.

BRIEF DESCRIPTION OF THE DRAWINGS [0012] In the drawings,

[0013] FIG. 1 is a side view of the present camera unit, according to an aspect of the disclosure;

[0014] FIG. 2 is a front view of the present unit, according to an aspect of the disclosure;

[0015] FIG. 3 is an alternative view of the front view of the camera unit, according to an aspect of the disclosure; and

[0016] FIG. 4 is a schematic diagram of the present system in use, according to an aspect of the disclosure.

DETAILED DESCRIPTION

[0017] Herein is disclosed a system for reading a tire code. Referring now to FIGS 1 -3, the system includes a camera unit 10 operable to capture a set of images of the sidewall of a tire at the location of the US DOT tire code. The set of images makes it possible to discern and read the tire code by a processor and then display it for a user to read. [0018] The term "camera unit" will be used herein to refer to a device that has a camera-like function but captures a set of images rather than just one at a time, and which set of images facilitates discernment of the information contained in the images. The term camera unit 10 will be further defined and characterized below.

[0019] The term "discern" means that the alphanumeric characters present in the set of images detected by their topographic features have significance as information and that the topographical features are parts of alphanumeric characters comprising the tire code have been captured in the images.

[0020] To improve the ability of the processor to read the tire code, multiple

images are taken with lighting directed at the tire from different angles within the camera to highlight topographic features of the tire sidewall, such as edges and flat surfaces, in the vicinity of the embossed or molded tire code.

[0021 ] Light external to the camera unit 10 is obscured by placing the camera unit 10 close to the tire sidewall at the location of the tire code, its window facing and centered on the tire code by the user. The processor may be physically separated from the camera unit 10 to reduce the size of the camera unit 10 and to provide the requisite computing power for reading the tire code. The processor is part of a server remote from the camera unit 10 but accessible via Wi-Fi or other wireless type of communication channel. Wi-fi is a facility that allows computers, smartphones, or other devices to connect to the Internet or to communicate with one another wirelessly within a particular area.

[0022] The camera unit 10 is a small, hand-held housing 14 that contains two side-by-side imaging devices 18, 22. Each imaging device 18, 22, has a field of view. The fields of view of these imaging devices 18, 22 overlap so that images captured by them can be presented as one image. Two (or more) imaging devices 18, 22, in a rectangular array enable the capturing of a combined, close- up image of the tire code. A tire code contains 12 (and sometimes fewer) alphanumeric characters in a sequence that fits within a thin, horizontal rectangle. Two smaller imaging devices 18, 22, in side-by-side relationship may capture a rectangular image more efficiently than a larger, square one. The camera unit 10 has a single transparent covering, such as glass or acrylic, over its front opening or window.

[0023] Imaging devices 18, 22, may have a single camera 34 with a lens 36 and an imager 38, as seen in FIGS. 1 and 2, or two cameras 18, 22, as shown in FIG. 3, which has a first lens 26 and a second lens 30, respectively, each of which has an imager 20 and 24, respectively.

[0024] The camera unit 10 may have an on-off button, perhaps a camera control button for the operator to verify that the image of the DOT code is visible and in focus, and an imaging sequence start button.

[0025] Also inside camera unit 10 are plural banks of lights, which may comprise a row of light-emitting diodes (LEDs). Banks of LEDs 38, 42, 46, and 50, may arranged on each of the four sides of the housing 14 of camera unit 10 near window and oriented so as to direct lighting through the glass to the exterior of camera unit 10 at different angles. Combinations of one or more of these banks of LEDs proceed in a pre-determined sequence to illuminate the tire code. The camera covering is brought close to the tire code so that the camera walls enclose the tire code and help to exclude ambient light when the banks of lights begin to flash in sequence. A sequence may be a single bank of lights from each of the four sides for four images and then two or four more images with different combinations of light banks at adjacent corners lighted.

[0026] When the camera is activated one or more banks of LEDs begin to

illuminate the tire code and to create the first of a sequence of high contrast images of the tire code. In a preselected sequence, the banks turn on, a pair of images, one from each imaging device, are acquired, and then the banks turn off and a different bank or banks turn on for the next pair of images, and so forth. The set of image pairs, which may be six pair of images, is forwarded from the camera unit 10, e.g., via a wireless protocol, to a handheld computer or mobile device containing an application for receiving images via Wi-Fi from the camera unit 10 and thence forwarding same to a processor adapted to read the images and determine the tire code. [0027] Reading images can be done with any of several techniques such as standard optical character recognition software or proprietary software including neural network techniques.

[0028] Convolutional neural networks are multi-layer neural networks. Like many other neural networks, they are programmed with a version of a back- propagation algorithm. Where they differ from other neural networks is in the architecture of the software.

[0029] Convolutional neural networks are designed to recognize visual patterns directly from pixel images with minimal preprocessing. They can recognize patterns with extreme variability (such as handwritten characters) and are robust to distortions and simple geometric transformations. LeNet-5 is an example of a convolutional network designed for handwritten and machine-printed character recognition.

[0030] The set of images of the tire code of a tire 54 made by the camera unit 10 and received by a handheld or mobile device 58 are transmitted, e.g., via Wi-Fi, to a cloud-base server 60 (or other server) for reading the code. The code, once read, is displayed on the handheld or mobile device 58 as a printed code, and may be displayed by the handheld or mobile device along 58 with the image captured by the camera unit 10.

[0031 ] The file may be stored or entered into a database depending on use. It may be transmitted to the hand held or mobile device 58, which may be a tablet or other small computer, for the intended use of the file, whether that be inventory control at a manufacturing warehouse or to check tire recall information at a repair shop.

[0032] Those skilled in the art of tire storage, control and inventory management will appreciate that there are many variations and substitutions that may be made in the foregoing description of aspects of the disclosure without departing from its spirit and scope.

Claims

WHAT IS CLAIMED IS:

1 . A device for reading a tire code, said device comprising:

(a) a camera unit operable to capture images of a tire code;

(b) at least two lights operable with respect to said camera to

illuminate said tire code;

(c) a processor in communication with said camera unit and operable to received said images from said camera unit;

(d) a server connected to said processor and operable to read said images from said camera unit and produce said tire code from said images as output; and

(e) a display operable to display said tire code received from said server.

2. The device of claim 1 , wherein said camera unit includes two cameras.

3. The device of claim 1 , wherein said camera unit includes two cameras, and wherein said server is operable to receive images from said two cameras.

4. The device of claim 1 , wherein said camera unit includes two cameras, and wherein said two cameras have fields of view that overlap.

5. The device of claim 1 , wherein said camera unit has two cameras and two lenses

6. The device of claim 1 , wherein said camera unit has two cameras in side-by-side relationship.

7. A device for reading a tire code, said device comprising:

(a) a camera unit operable to capture images of a tire code; (b) plural banks of lights in said camera unit sequentially operable to light said tire code;

(c) a server in communication with said camera unit and operable to receive images from said camera unit;

(d) a processor connected to said server and operable to read

images of a tire code and produce an output;

(e) a transceiver operable to receive said output from said processor and transmit said output; and

(f) a display operable to display said output received from said

transceiver.

8. 8 The device of claim 7, wherein said camera unit is operable to

obtain plural images, each image of said plural images lighted by a different combination of banks of lights of said plural banks of lights.

9. The device of claim 7, wherein said cameras system includes one camera.

10. The device of claim 7, wherein said camera unit includes two cameras in a side-by-side arrangement.

1 1 . The device of claim 7, wherein said camera unit includes two cameras in side-by-side arrangement, said two cameras operable to acquire said images, each image of said images being acquired when at least one bank of lights of said plural banks of lights is lighted.

12. The device of claim 7, wherein said camera unit lights said bank of lights of said plural banks of lights in a preselected sequence.

13. A method for reading a tire code, said method comprising the steps of:

(a) directing a camera unit at a tire code; (b) obscuring extrinsic light from said tire code;

(c) shining light from said plural banks of lights at said tire code;

(d) making plural images of said tire code, an image of said plural image for each angle of said plural angles;

(e) forwarding said plural images to a server for storage;

(f) processing said plural images to detect said tire code in plural images; and

(g) forwarding said tire code to a display.

The method of claim 13, wherein said camera unit includes two cameras and wherein said method further comprises the step of making said plural images as a pair of images, one image of said pair being captured by a first camera of said two cameras and a second image of said pair being captures by said second camera of said two cameras.

The method of claim 13, wherein said camera unit is two cameras and wherein said method further comprises the steps of:

a) orienting said two cameras side-by-side with overlapping fields of view; and

(b) making said plural images as a pair of images, one image of said pair being made by a first camera of said two cameras and a second image of said pair being made by said second camera of said two cameras.

The method of claim 13, wherein said light is from plural banks of lights, and wherein said method comprises the step of using at least one bank of lights of said plural banks of lights for each image of said plural images.

17. The method of claim 13, wherein said light is from plural banks of lights, and wherein said method comprises the step of using a preselected sequence of at least one bank of lights of said plural banks of lights for each image of said plural images.

18. The method of claim 13, wherein said light is from plural banks of lights, with at least one bank of lights of said plural banks of lights being used in a pre-selected sequence to obtain each image of said plural images.

19. The method of claim 13, wherein said camera unit includes two side- by-side cameras with overlapping fields of view, and wherein said method further comprises the step of obtaining plural pairs of images, a first image of a first pair of said plural pairs of images by a first camera of said two side-by-side camera and a second image of said first pair of said plural images by a second camera of said two side-by- side cameras.

20. The method of claim 19, wherein said light is plural banks of lights and wherein said method further comprises the step of selecting a different bank of lights for each pair of images.