Technique for embedding a code in an audio signal and for detecting the embedded code

• US 5,945,932 A
• Filed: 10/30/1997
• Issued: 08/31/1999
• Est. Priority Date: 10/30/1997
• Status: Expired due to Fees

• Current Assignee: Audiotrack Corporation
• Original Assignee: Audiotrack Corporation
• Inventors: Smith, Gregory L., Wiltshire, John H. D., Nolan, Marvin C.
• Primary Examiner(s): YOUNG, BRIAN K
• Application Number:
  US08/961,297
• Time in Patent Office: 670 Days
• Field of Search: 341/51, 341/94, 341/54, 341/55, 375/316, 375/377
• US Class Current: 341/51
• CPC Class Codes: G10L 19/018 Audio watermarking, i.e. embedding inaudible data in the audio signal
• CPC Class Codes: G11B 20/00891 embedded in audio data
• CPC Class Codes: H04B 1/665 using psychoacoustic properties of the ear, e.g. masking effect

Abstract

A code is embedded into an audio product so as to be processed therewith for recording and/or broadcast and yet be reliably detected while remaining inaudible to human perception. The code is represented by symbols formed from an impulse function having its energy within a specified frequency range. The audio product is analyzed to find segments which can mask the code based on tonality and a minimum signal energy. When the audio product with an embedded code is detected, decoding thereof involves finding candidate code signals which are checked against preset criteria. In particular, each symbol is made of at least two impulse functions with a preset spacing therebetween.

Claims

1. A method for embedding a digital code in a digitized audio product, comprising the steps of:

filtering the digitized audio product to a frequency band of interest;

determining a tonality indication for each of a plurality of segments of the filtered audio product which indicates the extent to which power is distributed uniformly for frequencies in at least a portion of said band of interest; and

inserting at least a portion of the digital code into a particular segment from said plurality of segments only if said tonality indication indicates a relatively uniform power distribution in said particular segment.

2. The method of claim 1, wherein said inserting step is performed only if at least one of the segments immediately before and immediately after said particular segment also has a tonality indication which indicates a relatively uniform power distribution.

3. The method of claim 1, wherein all of said plurality of segments have a uniform duration.

4. The method of claim 1, further comprising the step of determining the total power for at least said particular segment, and performing said inserting step only if said total power is above a predetermined threshold.

5. A method for embedding a digitized code in a digitized audio product, comprising the steps of:

filtering the digitized audio product to a frequency band of interest;

providing a coding signal derived from a band-limited impulse function with a waveform having its energy confined to and evenly spread across at least a portion of said frequency band of interest;

deriving said digitized code from said coding signal; and

embedding said digitized code into said audio product.

6. The method of claim 5, wherein said frequency band of interest is approximately 0 to 5000 Hz, and said coding signal has energy spread across approximately 1500 to 4000 Hz.

7. The method of claim 5, wherein said digitized code is derived based on polarity of said coding signal.

8. The method of claim 5, wherein each bit of said digitized code is derived based on a plurality of said coding signals.

9. The method of claim 5, wherein said coding signal is derived from an ideal mathematical impulse function to which a steep bandpass filter is applied.

10. A method for providing a digitized code to be embedded in a digitized audio product, comprising the steps of:

providing said digitized code as a series of binary bits;

dividing said binary bits into groups, each group having a plurality of bits;

providing coding signals to represent said bits, respectively;

deriving a symbol from said coding signals for each of said groups, each symbol having a plurality of said coding signals with a preset spacing therebetween.

11. The method of claim 10, wherein each symbol has coding signals equal in number to bits in a group to which such symbol corresponds.

12. The method of claim 10, wherein the coding signals are identical to each other in shape, with one binary bit corresponding to a coding signal of one polarity, and the other binary bit corresponding to a signal of the other polarity.

13. The method of claim 10, wherein a symbol consists of two of said coding signals.

14. The method of claim 10, wherein a symbol consists of three of said coding signals with equal spacing between adjacent ones.

15. A method for encoding and decoding a digitized code embedded in a digitized audio product, comprising the steps of:

deriving said digitized code in a form of start, data and end symbol types, each symbol representing a plurality of bits, and each bit being associated with a coding signal of given polarity;

generating said start type of symbol to consist of a plurality of said coding signals all of which have the same designated polarity;

embedding said digitized code in said digitized audio product;

detecting said digitized code embedded in said audio product; and

decoding said detected digitized code by determining whether the polarity of the coding signals on said start type of symbol is said designated polarity and, if not, inverting the polarity of said coding signals in said data and end types of symbols.

16. A method for embedding a digitized code in a digitized audio product, comprising the steps of:

identifying segments of the digitized audio product into which the digitized code can be embedded based on predetermined criteria;

generating portions of the digitized code for insertion into said segments, respectively;

removing the digitized audio product within said identified segments except for a predetermined small percentage of amplitude to generate modified segments; and

inserting said portions of the digitized code into said modified segments, respectively.

17. A method for embedding a digitized code in a digitized audio product, comprising the steps of:

analyzing the digitized audio product to derive measured values for designated characteristics thereof;

locating segments of the digitized audio product, based on said derived measured values and a set of preselected parameters, into which the digitized code can be inserted so as to be masked;

inserting the digitized code into said located segments;

determining whether a degree of masking of the inserted digitized code meets a predetermined level and, if not, modifying values of at least one of said set of preselected parameters, and then performing said locating and inserting steps again with said modified values.

18. A method for embedding a digitized code in an audio product, comprising the steps of:

dividing said digitized code into preselected portions;

representing said portions by a plurality of coding symbols, respectively;

determining spacing of said coding symbols from each other to be used for embedding the digitized code within the audio product so that said spacing is greater than a predetermined minimum, and

inserting the coding symbols within the audio product based on said determined spacing.

19. The method of claim 18, wherein said predetermined minimum spacing for a present symbol is in relation to a location of a previous symbol.

20. The method of claim 18, wherein said predetermined minimum spacing for a present symbol is in relation to a location of the immediately preceeding previous two symbols.

21. The method of claim 18, wherein said spacing is derived to produce a random spacing.

22. A method for decoding an audio product into which a code of digitized coding signals has been embedded, comprising the steps of:

obtaining a digitized audio product;

comparing said digitized audio product with a template of a coding signal to identify candidate coding signals based on shape;

comparing pairs of sequential candidate coding signals with each other based on preselected characteristics to identify which ones constitute the coding signals; and

reconstructing said code from said coding signals identified by said comparing step.

23. The method of claim 22, wherein said comparing step compares amplitude.

24. The method of claim 22, wherein said comparing step compares a predetermined spacing with spacing between a sequential pair of said candidate coding signals.