Title:
METHOD FOR DETECTING AND READING A MATRIX CODE MARKED ON A GLASS SUBSTRATE
Document Type and Number:
WIPO Patent Application WO/2022/042961
Kind Code:
A1
Abstract:
The invention pertains to computer implemented methods for detecting and reading a barcode marked on a glass substrate. The method is able to efficiently detect a possible barcode in a raster image and extract the valuable features of the image to provide an abstract representation for the barcode which contains only the information required for its processing by decoding or converting algorithms. More precisely, the method according to the invention relies on an image processing of a raster image of a barcode through chained convolutional artificial neural networks in order to provide a versatile, highly adaptive, memory efficient, and reliable probabilistic approach to detect and read barcodes. Thanks to the probabilistic approach, it can detect and read, on a various kinds of glass substrates, tracking codes, i.e. barcodes, in various formats, i.e. varying in size, brightness, contrast and/or aspect ratio, including shear, slant or trapezoidal distortions, local defects and/or non-uniform features, from any raster image of the portion of surface of a glass substrates onto which a tracking code may be present. The method is able to recognize features of the tracking codes in images beyond image distortions, so that it adapts to various image recording conditions.
Inventors:
LAUDEREAU JEAN-BAPTISTE (FR)
NIER VINCENT (FR)
NIER VINCENT (FR)
Application Number:
PCT/EP2021/070592
Publication Date:
March 03, 2022
Filing Date:
July 22, 2021
Export Citation:
Assignee:
SAINT GOBAIN (FR)
International Classes:
G06K7/14
Domestic Patent References:
| WO2014128424A1 | 2014-08-28 | |||
| WO2015121549A1 | 2015-08-20 |
Foreign References:
| US9396377B2 | 2016-07-19 | |||
| US20190384954A1 | 2019-12-19 | |||
| RU2726185C1 | 2020-07-09 | |||
| US20170024593A1 | 2017-01-26 | |||
| CN109543486A | 2019-03-29 | |||
| CN106446750A | 2017-02-22 | |||
| US20190099892A1 | 2019-04-04 |
Other References:
J. REDMONS. DIVVALA ET AL.: "You only look once: Unified, real-time object detection", PROCEEDINGS OF THE IEEE CONFERENCE ON COMPUTER VISION AND PATTERN RECOGNITION, 2016
K. HEX. ZHANG ET AL.: "Deep residual learning for image recognition", PROCEEDINGS OF THE IEEE CONFERENCE ON COMPUTER VISION AND PATTERN RECOGNITION, 2016
T.-Y. LINP. GOYAL ET AL.: "Focal loss for dense object detection", PROCEEDINGS OF THE IEEE INTERNATIONAL CONFERENCE ON COMPUTER VISION, 2017
O. RONNEBERGERP. FISCHER ET AL.: "U-net: Convolutional networks for biomedical image segmentation", INTERNATIONAL CONFERENCE ON MEDICAL IMAGE COMPUTING AND COMPUTER-ASSISTED INTERVENTION, 2015
K. HEX. ZHANG ET AL.: "Deep residual learning for image recognition", PROCEEDINGS OF THE IEEE CONFERENCE ON COMPUTER VISION AND PATTERN RECOGNITION, 2016
T.-Y. LINP. GOYAL ET AL.: "Focal loss for dense object detection", PROCEEDINGS OF THE IEEE INTERNATIONAL CONFERENCE ON COMPUTER VISION, 2017
O. RONNEBERGERP. FISCHER ET AL.: "U-net: Convolutional networks for biomedical image segmentation", INTERNATIONAL CONFERENCE ON MEDICAL IMAGE COMPUTING AND COMPUTER-ASSISTED INTERVENTION, 2015
Attorney, Agent or Firm:
SAINT-GOBAIN RECHERCHE (FR)
Download PDF: