Method, apparatus and system for recognising broadcast segments

Claim 2

The method of claim 1 wherein said signatures and parametized signals are derived from the video portion of said segments and said monitored signal.

Claim 3

The method of claim 1 wherein said signatures and parametized signals are derived from both the audio and video portions of said segments and said monitored signals.

Claim 4

The method of claim 2 wherein said signatures and said parametized signals are derived from the luminance of said video portion.

Claim 5

The method of claim 4 wherein each digital word of said signature and said parametized signals is derived by comparing the average luminances of selected areas of a frame of said segment to the average luminance of at least one reference area, each of said selected areas providing a bit of said digital word.

Claim 6

The method of claim 5 wherein said selected areas are asymmetrically spaced about the center of said frame.

Claim 7

The method of claim 5 wherein said at least one reference area is said entire frame.

Claim 8

The method of claim 5 wherein said at least one reference area is, for each of said selected areas, a different preselected area of said frame.

Claim 9

The method of claim 5 wherein said at least one reference area is, for each of said selected areas, the corresponding area of a previous frame.

Claim 10

The method of claim 5 wherein said at least one reference area is, for each of said selected areas, a different preselected area of a previous frame.

Claim 11

The method of claim 3 wherein said signatures and said parametized signals are derived at least in part from the frequency spectrum of said audio portion.

Claim 12

12 . The method of claim 11 wherein each digital word of said signatures and said paramatized signals is derived by comparing selected frequency bands of said audio portion to at least one reference band, each of said selected bands providing a bit of said digital word.

Claim 13

The method of claim 1 further comprising reducing the amount of data on which said searching step is performed.

Claim 14

The method of claim 13 wherein said data reduction step comprises associating the digital word representing a designated frame of each of said segments with the signature of said segment, said association being nonexclusive, said searching step comprising, for each digital word in said parametized monitored signal, searching said library for any stored signature for which said digital word represents said designated frame.

Claim 16

The method of claim 15 wherein said signatures and said parametized signals are derived from the video portion of said segments and said monitored signal.

Claim 17

The method of claim 15 wherein said signatures and said parametized signals are derived from both the audio and video portions of said segments and said monitored signals.

Claim 18

The method of claim 16 wherein said signatures and said parametized signals are derived from the luminance of said video portion.

Claim 19

The method of claim 18 wherein each digital word of said signature and said parametized signals is derived by comparing the average luminances of selected areas of a frame of said segment to the average luminance of at least one reference area, each of said selected areas providing a bit of said digital word.

Claim 20

The method of claim 19 wherein said selected areas are asymmetrically spaced about the center of said frame.

Claim 21

The method of claim 19 wherein said at least one reference area is said entire frame.

Claim 22

The method of claim 19 wherein said at least one reference area is, for each of said selected areas, a different preselected area of said frame.

Claim 23

The method of claim 19 wherein said at least one reference area is, for each of said selected areas, the corresponding area of a previous frame.

Claim 24

The method of claim 19 wherein said at least one reference area is, for each of said selected areas, a different preselected area of a previous frame.

Claim 25

The method of claim 17 wherein said signatures and said parametized signals are derived at least in part from the frequency spectrum of said audio portion.

Claim 26

The method of claim 25 wherein each digital word of said signatures and said parametized signals are derived by comparing selected frequency bands of said audio portion to at least one reference band, each of said selected bands providing one bit of said digital word.

Claim 27

Apparatus for continuous pattern recognition of broadcast segments, said apparatus comprising: means for constructing a digital signature from a known sample of a segment to be recognized by digitally parametizing said segments, selecting portions from among random frame locations throughout said parametized segment in accordance with a set of predefined rules to form said signature, and associating with said signature the frame locations of said portions; means for storing a library of signatures and associated frame locations, each signature in said library being identified with a particular segment to be recognized; means for monitoring a broadcast signal; means for digitally parametizing said monitored signal; means for searching said library for any signature potentially associated with each frame of said parametized monitored signal and, using the frame location information associated with said signature, for comparing each of said potentially associated stored signatures to the appropriate frames of said parametized monitored signal.

Claim 28

The apparatus of claim 27 wherein said constructing means operates on the video portion of said segment.

Claim 29

The apparatus of claim 27 wherein said constructing means operates on both the audio and video portions of said segment.

Claim 30

Apparatus for continuous pattern recognition of broadcast segments, said apparatus comprising: means for constructing a digital signature from a known sample of a segment to be identified by digitally parametizing said segment to form a string of digital words, each of said words representing one frame of said segment, identifying a word representing a designated frame in said segment, the association with said segment of the word representing said designated frame being nonexclusive, selecting a plurality of additional portions from among random frame locations throughout said segment in accordance with a set of predefined rules, and storing in a library of signatures the digital words representing said designated frame and said additional frames along with the offsets of said additional frames relative to said designated frame; means for monitoring a broadcast signal; means for digitally parametizing said monitored signal; and means for searching said library for any signature for which any monitored digital word represents said designated frame and comparing the additional words of any said signature with the words of said parametized monitored signal at said offsets relative to said monitored digital word.

Claim 31

The apparatus of Claim 30 wherein said constructing means operates on the video portion of said segment.

Claim 32

The apparatus of claim 30 wherein said constructing means operates on both the audio and video portions of said segment.

Claim 34

The system of claim 33 further comprising: a plurality of local sites, each of said sites located in a different geographic region for monitoring broadcast signals in said region; a central site; and a communications network linking said central site and said plurality of local sites; wherein: each of said local sites maintains a local library of segment signatures applicable to broadcast signals in its geographic region, each of said local sites having at least said storing means, said monitoring means, said searching means, said comparing means, and said detecting means; said central site has said constructing means and maintains a global library containing all of the information in all of said local libraries; and said classifying and identifying means comprises: means at said local site for generating compressed audio and video information, a temporary digital signature, and parametized monitored signal information for a potential unknown segment not found in the respective local library and for transmitting at least said parametized monitored signal information and said temporary digital signature for a potential unknown segment via said communications network to said central site; means at said central Bite for search ing and comparing signatures stored in said global library with said transmitted parametized monitored signal information; means at said central site for grouping together like potential unknown segments received by said central site from said local sites and not found in said global library; means at said central site for requesting at least one of said compressed audio and video information for at least one of said grouped potential unknown segments and for allowing an operator to play back said at least one of said audio and video information to classify said segment and to instruct said constructing means to automatically construct a signature for said segment; and means for adding said constructed signature to said global library and for transmitting it via said communications network to said local libraries.

Claim 35

The system of claim 34 wherein said grouping means comprises: means for comparing the temporary digital signatures of other potential unknown segments to a parametized potential unknown segment; for all potential unknown segments matching a given temporary digital signature, means for temporally aligning the parametized monitored signal information of said segments; and means for constructing a best fit to said aligned parametized monitored signal information.

Claim 36

The system of claim 34 further comprising means for transmitting from each of said local sites to said central site information concerning the occurrence of broadcast segments and means at said central site for logging said information.

Claim 37

The system of claim 36 further comprising means at said local site for storing the temporary digital signature of a potential unknown segment and for logging the occurrences of said potential unknown segment based on said stored temporary signature, and means for transmitting said log information to said central site after said central site has identified said potential unknown segment.

Claim 38

For use in a system for continuous pattern recognition of broadcast segments, said system including: . a plurality of local sites, each of said sites located in a different geographic region for monitoring broadcast signals in said region; a central site; and a communications network linking said central site and said plurality of local sites; each of said local sites maintaining a local library of segment signatures applicable to broadcast signals in its geographic region and having at least means for monitoring a broadcast signal, means for searching said local library for any signature potentially associated with each frame of said parametized monitored signal and for comparing said signatures to said parametized monitored signal, said means for detecting, in said monitored signal, the occurrence of artifacts characteristic of potential unknown segments to be recognized; and said central site maintaining a global library containing all of the information stored in all of said local libraries; the method of classi- tying and identifying said potental unknown segments, said method comprising: parametizing at one of said local sites said potential unknown segment; constructing at said local site a temporary digital signature for said potential unknown segment; generating at said local site compressed digitized audio and video information for said potential unknown segment; transmitting said parametized unknown segment and said temporary digital signature, and, as determined by said central site at least one of said compressed audio and video information from said local site to said central site; comparing the sigantures in said global library to said parametized unknown segment to determine if said unknown segment was previously received by another of said local sites and if so replacing said temporary signature at said local site with the appropriate global signature; grouping together like unknown segments received from several local sites; allowing an operator to play back said at least one of said compressed audio and video information to identify said unknown segments; constructing a new global signature for said identified signal; transmitting said new global signature to those of said local sites for which it is appropriate.

Claim 39

3 9 . The method of claim 38 wherein said grouping step comprises: comparing the temporary digital signatures of other potential unknown segments to a parametized potential unknown segment; for all potential unknown segments matched by a given temporary digital signature, temporally aligning the parametized signal information of said segments; and constructing a best fit to said aligned parametized signal information.

Claim 40

The method of claim 39 wherein said step of constructing a best fit comprises polling each bit position of such aligned parametized signal information and assigning to the corresponding bit position of said best fit the value of the majority of parametized segments in said bit location.

Claim 41

The method of claim 38 further comprising confirming said new global signature by testing said new global signature against each potential unknown segment in said group.

Claim 43

The method of claim 42 wherein: - said step of recording additional data comprises constructing a second signature for said variable portion; and said using step comprises identifying the variable portion of said broadcast segment with said second signature.

Claim 44

The method of claim 42 wherein: said step of recording additional data comprises recording at least one video frame of said variable portion; and said using step comprises performing a direct comparison of said at least one video frame and at least one frame of said monitored broadcast signal.

Claim 45

Apparatus for continuous pattern recognition of broadcast segments, each of said segments having a constant portion and a variable portion, said apparatus comprising: means for constructing a signature from a known sample of the constant portion of such a segment to be recognized; means for recording additional data from known samples of said variable portion; means for storing said signature andsaid additional data; means for associating with said stored signature an indication of the existence of said variable portion; means for monitoring a broadcast signal; means for identifying the constant portion of a broadcast segment in said monitored broadcast signal with said signature; and means for using said additional data to identify the variable portion of said broadcast segment.

Claim 46

The apparatus of claim 45 wherein: said recording means comprises means for constructing a second signature for said variable portion; and said using means comprises identifying the variable portion of said broadcast segment with said second signature.

Claim 47

The apparatus of claim 45 wherein: said recording means comprises means for recording at least one video frame of said variable portion; and said using means comprises means for performing a direct comparison of said at least one video frame and at least one frame of said monitored broadcast signal.