Localization of machine-readable indicia in digital capture systems
The present disclosures relates to finding or localizing machine readable indicia (e.g., a barcode or digital watermark) in imagery. One claim recites an apparatus comprising: memory for buffering blocks of image data, the image data having been captured with a camera and depicting a printed object; one or more processors programmed for: generating an edge orientation sensitive feature set from the image data; using a first trained classifier to determine whether the feature set includes data representing a barcode; and using N additional trained classifiers to determine an orientation angle associated with the barcode, wherein N comprises an integer greater than 3, and wherein the orientation angle is selected based on a probability metric. Of course, other claims and combinations are provided too.
1. A method comprising:
obtaining data representing captured imagery, the captured imagery depicting printed or displayed material;
using one or more programmed multi-core processors:
generating a feature set from the data representing captured imagery;
determining, using a first trained classifier, whether the feature set includes data representing a 1D-barcode; and
determining, using N additional trained classifiers, an orientation angle associated with the 1D-barcode, wherein N comprises an integer greater than 3.
2. The method of claim 1 in which at least two orientation angles associated with the 1D-barcode are determined using the N additional trained classifiers.
3. The method of claim 1 in which the N additional trained classifiers operate on a feature set centered at or around an image area associated with the 1D-barcode, in which the image area is a subset of the data representing captured imagery.
4. The method of claim 1 in which the first trained classifier is trained based on a binary decision of present or not.
5. The method of claim 1 in which the first trained classifier is trained based on a linear regressor.
6. An apparatus comprising:
memory for buffering image data, the image data having been captured with a camera and depicting a printed object;
one or more multi-core processors programmed for:
generating a feature set from buffered image data;
using a first trained classifier to determine whether the feature set includes data representing a barcode; and
using N additional trained classifiers to determine an orientation angle associated with the barcode, wherein N comprises an integer greater than 3, and wherein the orientation angle is determined based on a probability metric.
7. The apparatus of claim 6 in which at least two orientation angles associated with the barcode are determined using the N additional trained classifiers, and in which the barcode comprises a 1D-barcode.
8. The apparatus of claim 6 in which the N additional trained classifiers operate on a feature set centered at or around an image area associated with the barcode, in which the image area is a subset of the buffered image data.
9. The apparatus of claim 6 in which the first trained classifier is trained based on a binary decision of barcode present or not.
10. The apparatus of claim 6 in which the first trained classifier is trained based on a linear regressor.
11. An apparatus comprising:
electronic memory for buffering image data, the image data having been captured with a digital camera and depicting a printed or displayed object;
means for generating a feature set from the image data, the feature set being associated with edge orientation of the image data;
means for determining whether the feature set includes data representing a barcode, in which said means for determining whether the feature set includes data representing a barcode utilizes a first trained classifier; and
means for determining an orientation angle associated with the barcode, in which said means for determining an orientation angle utilizes N additional trained classifiers to, wherein N comprises an integer greater than 3, and wherein the orientation angle is determined based on a probability metric.
12. The apparatus of claim 11 in which at least two orientation angles associated with the barcode are determined using the N additional trained classifiers.
13. The apparatus of claim 11 in which the N additional trained classifiers operate on a feature set centered at or around an image area associated with the barcode, in which the image area is a subset of the image data, and in which the barcode comprises a 1-D barcode.
14. The apparatus of claim 11 in which the first trained classifier is trained based on a binary decision of barcode present or not.
15. The apparatus of claim 11 in which the first trained classifier is trained based on a linear regressor.
16. The apparatus of claim 11 further comprising means for generating a 2-dimensional heat-map including different colors corresponding to different probabilities of image areas likely including a barcode, the heat-map including a representation of at least a portion of the depicted printed or displayed object; and means for displaying the heat-map.
17. The apparatus of claim 6 in which the feature set comprises an association with edge origination of the image data.
18. The apparatus of claim 17 in which the feature set comprises an edge orientation sensitive feature set.
19. The method of claim 1 in which the feature set comprises an association with edge orientation of the image data.
20. The method of claim 19 in which the feature set comprises an edge orientation sensitive feature set.