Robust spectral encoding and decoding methods

• US 10,192,560 B2
• Filed: 09/25/2017
• Issued: 01/29/2019
• Est. Priority Date: 05/22/2007
• Status: Active Grant

• Current Assignee: Digimarc Corporation
• Original Assignee: Digimarc Corporation
• Inventors: Sharma, Ravi K., Alattar, Adnan M.
• Primary Examiner(s): Sharma, Neeraj
• Application Number:
  US15/714,642
• Publication Number:
  US 20180130477A1
• Time in Patent Office: 491 Days
• Field of Search: None
• CPC Class Codes: G10L 15/08 Speech classification or search
• CPC Class Codes: G10L 19/00 Speech or audio signals analysis-synthesis techniques for redundancy reduction, e.g. in vocoders; Coding or decoding of speech or audio signals, using source filter models or psychoacoustic analysis in musical instruments G10H
• CPC Class Codes: G10L 19/018 Audio watermarking, i.e. embedding inaudible data in the audio signal
• CPC Class Codes: G10L 19/022 Blocking, i.e. grouping of samples in time; Choice of analysis windows; Overlap factoring
• CPC Class Codes: G10L 19/12 the excitation function being a code excitation, e.g. in code excited linear prediction [CELP] vocoders
• CPC Class Codes: G10L 19/265 Pre-filtering, e.g. high frequency emphasis prior to encoding
• CPC Class Codes: G10L 21/055 for synchronising with other signals, e.g. video signals

Abstract

Spectral encoding methods are more robust when used with improved weak signal detection and synchronizations methods. Further robustness gains are achieved by using informed embedding, error correction and embedding protocols that enable signal to noise enhancements by folding and pre-filtering the received signal.

Claims

1. A device for decoding a digital watermark embedded in an audio signal, wherein the digital watermark is embedded in the audio signal by adjusting signal values, the device comprising:

a memory in which is stored blocks of the audio signal;

a processor in communication with the memory to obtain blocks of the audio signal, the processor configured with instructions to;

perform an initial synchronization of the digital watermark, distorted due to time scale change, by converting blocks of the audio to frequency domain data, pre-filtering the frequency domain data to produce first pre-filtered blocks, summing the first pre-filtered blocks to produce a first accumulated block, the first pre-filtered blocks being selected over a sufficiently long block of audio such that plural instances of watermark signal representing the same code at different time locations are accumulated, and correlating the first accumulated block with a pattern to detect a time scale of an embedded code signal; and

perform decoding of variable code data of the digital watermark at the detected time scale by correlating phase of the audio signal at the detected shift with a data signal pattern, wherein the processor is configured with instructions to decode the variable code data by correlating phase of the audio signal with a synchronization code signal to identify a start position of the variable code, and correlating phase of the audio signal with data code signals to detect data code signals encoded as the variable code data of the digital watermark.

2. The device of claim 1 wherein the processor is configured with instructions to perform a least squares method to approximate time scale changes of the data embedded in the audio signal.

3. The device of claim 1 wherein the processor is configured with instructions to perform a soft correlation between a pattern signal at selected frequencies and the audio signal to detect the shift of the embedded code signal.

7. The device of claim 2 wherein the processor is configured with instructions to determine a least squares mapping between peak locations of the frequency domain data in the first accumulated block and code signal locations.

8. The device of claim 7 wherein first accumulated block comprises frequency domain data summed from plural sync blocks.

9. The device of claim 1 wherein the processor is configured with instructions to perform a soft correlation between a pattern signal and the first accumulated block to detect variable code data.

4. A tangible non-transitory, computer readable medium, on which is stored instructions, which, when executed by a computer, cause the computer to decode a digital watermark embedded in an audio signal, wherein the digital watermark is embedded in the audio signal by adjusting signal values, the instructions, which when executed by the computer, cause the computer to:

perform an initial synchronization of the digital watermark, distorted due to time scale change by converting blocks of the audio to frequency domain data, pre-filtering the frequency domain to produce first pre-filtered blocks, summing the first pre-filtered blocks to produce a first accumulated block, the first pre-filtered blocks being selected over a sufficiently long block of audio such that plural instances of watermark signal representing the same code at different time locations are accumulated, and correlating the first accumulated block with a pattern to detect a time scale of an embedded code signal; and

perform decoding of variable code data of the digital watermark at the detected time scale by correlating phase of the audio signal at the detected shift with a data signal pattern, wherein the instructions on the medium cause the computer to decode the variable code data by correlating phase of the audio signal with a synchronization code signal to identify a start position of the variable code, and correlating phase of the audio signal with data code signals to detect data code signals encoded as the variable code data of the digital watermark.

5. The non-transitory computer readable medium of claim 4, comprising instructions, which when executed by the computer, perform a least squares method to approximate time scale changes of the data embedded in the audio signal.

6. The non-transitory computer readable medium of claim 4, comprising instructions, which when executed by the computer, perform a soft correlation between a pattern signal at selected frequencies and the audio signal to detect the shift of the embedded code signal.

10. The non-transitory computer readable medium of claim 5 comprising instructions, which when executed by a computer, cause the computer to determine a least squares mapping between peak locations of the frequency domain data in the first accumulated block and code signal locations.

11. The non-transitory computer readable medium of claim 10 wherein first accumulated block comprises frequency domain data of plural sync blocks.

12. The non-transitory computer readable medium of claim 4 comprising instructions, which when executed by a computer, cause the computer to perform a soft correlation between a pattern signal and the first accumulated block to detect variable code data.

13. A method for decoding a digital watermark embedded in an audio signal, the method comprising:

converting blocks of the audio signal to frequency domain data;

pre-filtering the frequency domain data to produce first pre-filtered blocks;

summing the first pre-filtered blocks to produce a first accumulated block, the first pre-filtered blocks being selected over a sufficiently long block of audio such that plural instances of watermark signal representing the same code at different time locations are accumulated;

detecting signal peaks in the first accumulated block;

determining a time scale change of the digital watermark based on locations of the signal peaks; and

decoding variable code data encoded in the digital watermark by correlating phase of the audio signal at the determined time scale change with a data signal pattern, the decoding comprising correlating phase of the audio signal with a synchronization code signal to identify a start position of the variable code, and correlating phase of the audio signal with data code signals to detect data code signals encoded as the variable code data of the digital watermark.

14. The method of claim 13 wherein the determining includes correspondence between the signal peaks and code signal elements, and determining a transform that maps code signal elements to corresponding signal peaks through a least squares estimation.

15. The method of claim 13 further including correlating plural code signal patterns corresponding to data signal elements with the accumulated data of the first block to detect variable code signals within the audio signal.