Label capture using optical pattern recognition and barcode semantics
AssigneeScandit AG
InventorsChristian Kündig, Oliver Akermann, Bernd Schoner, Luc Humair, Luis Haug, Marco Maggipinto
Life of this patent
Estimating the term…
Abstract
One or more images of a real scene are analyzed to identify a plurality of text fields and one or more barcodes. A plurality of edges are defined, wherein each edge of the plurality of edges links a text field of the plurality of text fields with a barcode of the one or more barcodes. The plurality of edges are used to match a text field with a barcode to classify the barcode.
Claims20
A system for image analysis of optical patterns in a real scene, the system comprising:
- a camera;
- and
- one or more computer-readable media comprising instructions that, when executed by one or more processors, perform the following steps:
- receiving one or more images of a scene acquired by the camera;
- performing image processing on an image of the one or more images to computationally identify boundaries for a plurality of text fields within the
- image and for one or more barcodes within the image;
- constructing a data graph that represents spatial relationships between the plurality of text fields and the one or more barcodes, wherein the data graph comprises a plurality of edges, wherein each edge of the plurality of edges links a text field node associated with a text field of the plurality of text fields with a barcode node associated with a barcode of the one or more barcodes, and wherein the data graph is constructed based on the boundaries;
- and
- analyzing the data graph to identify a correct match between a text field of the plurality of text fields and a corresponding barcode of the one or more barcodes, and in response to identifying the correct match, automatically decoding the corresponding barcode to extract its data.
A method for image analysis of optical patterns in a real scene, the method comprising:
- receiving one or more images of a scene acquired by a camera;
- performing image processing on an image of the one or more images to computationally identify boundaries for a plurality of text fields within the
- image and for one or more barcodes within the image;
- and
- constructing a data graph that represents spatial relationships between the plurality of text fields and the one or more barcodes, wherein the data graph comprises a plurality of edges, wherein each edge of the plurality of edges links a text field node associated with a text field of the plurality of text fields with a barcode node associated with a barcode of the one or more barcodes, and wherein the data graph is constructed based on the boundaries;
- and
- analyzing the data graph to identify a correct match between a text field of the plurality of text fields and a corresponding barcode of the one or more barcodes, and in response to identifying the correct match, automatically decoding the corresponding barcode to extract its data.
A non-transitory computer-readable storage medium comprising instructions that, when executed by one or more processors, perform the following steps for image analysis of optical patterns in a real scene:
- receiving one or more images of a scene acquired by a camera;
- performing image processing on an image of the one or more images to computationally identify boundaries for a plurality of text fields within the
- image and for one or more barcodes within the image;
- constructing a data graph that represents spatial relationships between the plurality of text fields and the one or more barcodes, wherein the data graph comprises a plurality of edges, wherein each edge of the plurality of edges links a text field node associated with a text field of the plurality of text fields with a barcode node associated with a barcode of the one or more barcodes, and wherein the data graph is constructed based on the boundaries;
- and
- analyzing the data graph to identify a correct match between a text field of the plurality of text fields and a corresponding barcode of the one or more barcodes, and in response to identifying the correct match, automatically decoding the corresponding barcode to extract its data.
Dependencies are inferred from the claim text.