Decoding machine-readable optical codes with aesthetic component
Techniques are provided for decoding machine-readable optical codes that have an aesthetic component that is integrated into the codes themselves. In this manner, the machine-readable optical codes can be designed to be aesthetically pleasing and/or can convey information to human viewers, and can even be disguised so that they do not appear to be machine-readable optical codes at all. Such information can be (but need not be) distinct from the information encoded for reading by a machine, even when the information is integrated into the code itself. The techniques described herein can be applied to any type of machine-readable optical code.
1. A computer-implemented method for detecting a fiducial containing a machine-readable optical code, comprising:
at an image capture device, capturing an input image;
at a processor, determining a local scale for each of a plurality of locations within the input image, to generate a locally averaged image;
at the processor, comparing the locally averaged image at each location to a threshold;
at the processor, generating a binary image based on the comparison;
at the processor, identifying connected components within the binary image;
at the processor, tracing region boundaries within the connected components;
at the processor, identifying quadrilaterals from the traced region boundaries;
at the processor, filtering the identified quadrilaterals to generate a set of at least one quadrilateral; and
at an output device, outputting the set of at least one quadrilateral as a set of at least one detected fiducial containing a machine-readable optical code.
2. The computer-implemented method of claim 1 , wherein the machine-readable optical code comprises an aesthetic component.
3. The computer-implemented method of claim 1 , further comprising, prior to tracing region boundaries, filtering the connected components to determine which regions are above a threshold metric to represent meaningful data.
4. The computer-implemented method of claim 3 , wherein the threshold metric comprises at least one of size and solidity.
5. The computer-implemented method of claim 1 , further comprising, prior to outputting the set of at least one quadrilateral, refining the at least one quadrilateral using image-based alignment.
6. A computer-implemented method for detecting a fiducial containing a machine-readable optical code, comprising:
at an image capture device, capturing an input image;
at a processor, generating a binary image from the input image;
at the processor, identifying connected components within the binary image;
at the processor, tracing region boundaries within the connected components;
at the processor, identifying quadrilaterals from the traced region boundaries;
at the processor, filtering the identified quadrilaterals to generate a set of at least one quadrilateral;
at an output device, outputting the set of at least one quadrilateral as a set of at least one detected fiducial containing a machine-readable optical code;
at the processor, obtaining a canonical fiducial containing canonical probe locations;
at the processor, determining a perspective transformation of the at least one detected fiducial as compared with the canonical fiducial;
at the processor, applying the perspective transformation to the canonical probe locations to obtain transformed probe locations;
at the processor, determining at least one probe value for each transformed probe location;
at the processor, determining binary probe values from the determined probe values;
at the processor, comparing the determined binary probe values with a code library to identify a matched code; and
at an output device, outputting the matched code.
7. The computer-implemented method of claim 6 , wherein determining at least one probe value for each transformed probe location comprises, for each transformed probe location:
determining a plurality of probe values; and
averaging the probe values.
8. The computer-implemented method of claim 6 , wherein comparing the determined binary probe values with a code library to identify a matched code comprises determining an orientation based on at least one orientation marker.
9. A system for detecting a fiducial containing a machine-readable optical code, comprising:
an image capture device, configured to capture an input image;
a hardware processor, communicatively coupled to the image capture device, configured to perform the steps of:
determining a local scale for each of a plurality of locations within the input image, to generate a locally averaged image;
at each location, comparing the locally averaged image to a threshold;
generating a binary image based on the comparison;
identifying connected components within the binary image;
tracing region boundaries within the connected components;
identifying quadrilaterals from the traced region boundaries; and
filtering the identified quadrilaterals to generate a set of at least one quadrilateral; and
an output device, communicatively coupled to the hardware processor, configured to output the set of at least one quadrilateral as a set of at least one detected fiducial containing a machine-readable optical code.
10. The system of claim 9 , wherein the machine-readable optical code comprises an aesthetic component.
11. The system of claim 9 , wherein the hardware processor is further configured to, prior to tracing region boundaries, filter the connected components to determine which regions are above a threshold metric to represent meaningful data.
12. The system of claim 11 , wherein the threshold metric comprises at least one of size and solidity.
13. The system of claim 9 , wherein the hardware processor is further configured to, prior to the output device outputting the set of at least one quadrilateral, refine the at least one quadrilateral using image-based alignment.
14. A system for detecting a fiducial containing a machine-readable optical code, comprising:
an image capture device, configured to capture an input image;
a hardware processor, communicatively coupled to the image capture device, configured to perform the steps of:
generating a binary image from the input image;
identifying connected components within the binary image;
tracing region boundaries within the connected components;
identifying quadrilaterals from the traced region boundaries;
filtering the identified quadrilaterals to generate a set of at least one quadrilateral;
obtaining a canonical fiducial containing canonical probe locations;
determining a perspective transformation of the detected fiducial as compared with the canonical fiducial;
applying the perspective transformation to the canonical probe locations to obtain transformed probe locations;
determining at least one probe value for each transformed probe location;
determining binary probe values from the determined probe values; and
comparing the determined binary probe values with a code library to identify a matched code; and
an output device, communicatively coupled to the hardware processor, configured to output the set of at least one quadrilateral as a set of at least one detected fiducial containing a machine-readable optical code, and further configured to output the matched code.
15. The system of claim 14 , wherein the hardware processor determining at least one probe value for each transformed probe location comprises, for each transformed probe location:
determining a plurality of probe values; and
averaging the probe values.
16. The system of claim 14 , wherein the hardware processor comparing the determined binary probe values with a code library to identify a matched code comprises determining an orientation based on at least one orientation marker.
17. A system for decoding a machine-readable optical code contained in a detected fiducial, comprising:
an image capture device, configured to capture an input image;
a hardware processor, communicatively coupled to the image capture device, configured to perform the steps of:
detecting a fiducial within the input image, the fiducial comprising a machine-readable optical code;
obtaining a canonical fiducial containing canonical probe locations;
determining a perspective transformation of the detected fiducial as compared with the canonical fiducial;
applying the perspective transformation to the canonical probe locations to obtain transformed probe locations;
determining at least one probe value for each transformed probe location;
determining binary probe values from the determined probe values; and
comparing the determined binary probe values with a code library to identify a matched code; and
an output device, communicatively coupled to the hardware processor, configured to output the matched code.
18. The system of claim 17 , wherein the hardware processor determining at least one probe value for each transformed probe location comprises, for each transformed probe location:
determining a plurality of probe values; and
averaging the probe values.
19. The system of claim 17 , wherein the hardware processor comparing the determined binary probe values with a code library to identify a matched code comprises determining an orientation based on at least one orientation marker.
20. A non-transitory computer-readable medium for detecting a fiducial containing a machine-readable optical code, comprising instructions stored thereon, that when executed on a processor, perform the steps of:
causing an image capture device to capture an input image;
determining a local scale for each of a plurality of locations within the image, to generate a locally averaged image;
at each location, comparing the locally averaged image to a threshold;
generating a binary image based on the comparison;
identifying connected components within the binary image;
tracing region boundaries within the connected components;
identifying quadrilaterals from the traced region boundaries;
filtering the identified quadrilaterals to generate a set of at least one quadrilateral; and
causing an output device to output the set of at least one quadrilateral as a set of at least one fiducial containing a machine-readable optical code.
21. The non-transitory computer-readable medium of claim 20 , wherein the machine-readable optical code comprises an aesthetic component.
22. The non-transitory computer-readable medium of claim 20 , further comprising, prior to tracing region boundaries, filtering the connected components to determine which regions are above a threshold metric to represent meaningful data.
23. The non-transitory computer-readable medium of claim 22 , wherein the threshold metric comprises at least one of size and solidity.
24. The non-transitory computer-readable medium of claim 20 , wherein the non-transitory computer-readable medium comprises further instructions stored thereon, that when executed on a processor, perform the step of, prior to outputting the set of at least one quadrilateral, refining the at least one quadrilateral using image-based alignment.
25. A non-transitory computer-readable medium for detecting a fiducial containing a machine-readable optical code, comprising instructions stored thereon, that when executed on a processor, perform the steps of:
causing an image capture device to capture an input image;
generating a binary image from the input image;
identifying connected components within the binary image;
tracing region boundaries within the connected components;
identifying quadrilaterals from the traced region boundaries;
filtering the identified quadrilaterals to generate a set of at least one quadrilateral;
causing an output device to output the set of at least one quadrilateral as a set of at least one fiducial containing a machine-readable optical code;
obtaining a canonical fiducial containing canonical probe locations;
determining a perspective transformation of the detected fiducial as compared with the canonical fiducial;
applying the perspective transformation to the canonical probe locations to obtain transformed probe locations;
determining at least one probe value for each transformed probe location;
determining binary probe values from the determined probe values;
comparing the determined binary probe values with a code library to identify a matched code; and
causing the output device to output the matched code.
26. The non-transitory computer-readable medium of claim 25 , wherein determining at least one probe value for each transformed probe location comprises, for each transformed probe location:
determining a plurality of probe values; and
averaging the probe values.
27. The non-transitory computer-readable medium of claim 25 , wherein comparing the determined binary probe values with a code library to identify a matched code comprises determining an orientation based on at least one orientation marker.