Method and apparatus for classifying patterns of television programs and commercials based on discerning of broadcast audio and video signals

• US 5,343,251 A
• Filed: 05/13/1993
• Issued: 08/30/1994
• Est. Priority Date: 05/13/1993
• Status: Expired due to Fees

• Current Assignee: Pareto Partners Ltd. (VAST.com)
• Original Assignee: Pareto Partners Ltd. (VAST.com)
• Inventors: Nafeh, John
• Primary Examiner(s): Groody, James J.
• Assistant Examiner(s): Metjahic, Safet
• Application Number:
  US08/062,458
• Time in Patent Office: 474 Days
• Field of Search: 358/139, 358/10, 358/185, 358/160, 358/335, 358/311, 358/312, 358/310, 358/908, 360/27, 360/14.1, 360/33.1, 360/74.1, 360/71, 360/69, 455/67, 395/22, 395/21, 348/907, 348/722, 348/571, 348/5, 348/180
• US Class Current: 348/571
• CPC Class Codes: G11B 15/026 by using processor, e.g. microcomputer
• CPC Class Codes: G11B 15/087 by sensing recorded signals
• CPC Class Codes: G11B 15/1875 adaptations for special effects or editing signal processing or indexing therefor G11B27/00
• CPC Class Codes: H04H 2201/90 characterised by the use of signatures
• CPC Class Codes: H04H 60/372 Programme
• CPC Class Codes: H04H 60/375 Commercial
• CPC Class Codes: H04H 60/58 of audio determination or detection of speech characteristics in general G10L25/00; speech recognition in general G10L15/00
• CPC Class Codes: H04H 60/59 of video recognising characters or patterns in general G06F18/00, G06V20/00
• CPC Class Codes: H04N 17/06 for recorders
• CPC Class Codes: H04N 5/765 Interface circuits between an apparatus for recording and another apparatus
• CPC Class Codes: H04N 5/782 on tape
• CPC Class Codes: H04N 5/9201 involving the multiplexing of an additional signal and the video signal
• CPC Class Codes: Y10S 358/908 Pause control, i.e. "commercial killers"

Abstract

A method and apparatus for classifying patterns of television programs and commercials, based on learning and discerning of broadcast audio and video signals, wherein the signals might be incomplete and the number of discerning features is large. The apparatus uses a discerner device containing feature extraction, an artificial neural network and control mechanisms. The discerner device classifies signal patterns into classes and stores, records or displays them accordingly on different storage, recording or display devices using the appropriate control mechanisms. The method is operable for classifying many types of signal patterns, but most importantly, those patterns of programs and commercials of television broadcast audio and video signals.

Claims

1. Apparatus for processing electromagnetic signals capable of being separated into a first group and a second group comprising:

means for receiving said signals as a signal stream;

means coupled with the receiving means for forming the signal stream into segments which can be represented as parameters, with the groups being mixed together;

means coupled with said forming means for storing the signal parameters after they have been formed;

means coupled with said storing means for classifying the segments into the first and second groups, said classifying means being operable to generate at least one decisional index representative of at least one signal segment and its parameters of at least one of the said groups of signals;

means for storing past decisions and for coupling the last-mentioned storing means to the classifying means to allow such past decisions to be included in the classifying decisions to determine which signal groups the signal segments belong to; and

a signal controller coupled with the classifying means for receiving and being responsive to a decisional index therefrom, said controller being operable to control a signal sensitive structure adapted for directing and transforming said groups of signals to respective end use locations.

2. Apparatus for processing signals as set forth in claim 1, wherein said first group of signals includes program signals and the second group of signals includes commercial signals.

3. Apparatus as set forth in claim 1, wherein the means for classifying the segments includes a neural network.

4. Apparatus as set forth in claim 1, wherein is included means defining a transition indicator for coupling the storage means to the classifying means, said transition indicators being operable for actuating the classifying means to cause a decision to be made by the classifying means as to the classification of at least one signal segment, relative to said first and second groups.

5. Apparatus as set forth in claim 1, wherein is included a transition indicator means for coupling the storage means to the classifying means, said transition indicator being operable for actuating the classifying means to cause decision to be made by the classifying of each segment relative to the first and second groups.

6. Apparatus for processing electromagnetic signals representing program and commercial signals comprising:

means adapted to be coupled to a source of said signals for receiving the signals as a signal stream;

means coupled with the receiving means for forming the signal stream into segments which can be represented as parameters, with the groups being mixed together;

means coupled with the forming means for storing the signal parameters after they have been formed;

a neural network coupled with said storing means and including classifying means operable to generate at least one decisional index representative of at least one signal segment and its parameters of at least one of said groups of signals;

means defining a transition indicator for coupling the storage means to the classifying means, said transition indicator being operable for actuating the neural network to cause a decision to be made by the neural network as to the classification of at least one signal segment, relative to said program and commercial signals;

means for storing past decisions;

feedback means for coupling the last-mentioned storing means to the classifying means to allow such past decisions to be included in the classifying decisions to determine which of said signal groups the signal segments belong to; and

a signal controller coupled with the neural network for receiving and being responsive to a decisional index from the neural network, said controller being operable to control a signal sensitive structure adapted for directing and transforming said groups of signals to respective end use locations.

7. Apparatus as set forth in claim 6, wherein said electromagnetic signals include audio signals, chrominance signals, luminance signals, vertical integral time code (VITC) signal, a close-captioned (CC) code signal, a color carrier jitter (CCJ) signal, and a signal to noise ratio (SNR) signal.

8. Apparatus as set forth in claim 6, wherein said audio signals are stereo signals.

9. Apparatus as set forth in claim 7, wherein the audio signals are divided into a number of bands, each band being the source of signals from which said parameters are determined.

10. A method for processing electromagnetic signals capable of being separated into a first group and a second group comprising:

receiving said signals at a location as a signal stream;

forming the signal stream into segments which can be represented as parameters, with the groups being mixed together;

storing the signal parameters after they have been formed;

classifying at least one segment into the first and second groups;

generating as a function of said classifying step at least one decisional index representative of at least one signal segment and its parameters of at least one of the said groups of signals;

storing past decisions; and

allowing such past decisions to be included in the classifying decisions to determine which signal groups the signal segments belong to and in response to a decisional index, controlling a signal-sensitive structure adapted for directing and transforming said groups of signals to respective end use locations.

11. A method as set forth in claim 10, wherein said first group of signals includes program signals and the second group of signals includes commercial signals.

12. A method as set forth in claim 10, wherein the classifying step includes operating at least one neural network.

13. A method as set forth in claim 10, wherein is included the step of sensing a transition in at least one signal segment from one said group to another group;

and

actuating the classifying means to cause a decision to be made by the classifying means as to the classification of at least one signal segment, relative to said first and second groups.

14. A method for processing electromagnetic signals representing program and commercial television signals comprising:

receiving the signals as a signal stream at a first location;

forming the signal stream into segments which can be represented as parameters, with the groups being mixed together;

storing the signal parameters after they have been formed;

operating a neural network after the storing step to classify at least one signal segment and to generate at least one decisional index representative of at least one signal segment and its parameters of at least one of said groups of signals;

sensing a transition from one group to the other group by actuating the neural network to cause a decision to be made by the neural network as to the classification of at least one signal segment, relative to said program and commercial signals;

storing past decisions;

allowing such past decisions to be included in the classifying decisions to determine which of said signal groups the signal segments belong to; and

controlling a signal-sensitive structure adapted for directing and transforming said groups of signals to respective end use locations.

15. A method as set forth in claim 14, wherein said electromagnetic signals include audio signals, chrominance signals, luminance signals, vertical integral time code (VITC) signal, a close-captioned (CC) code signal, a color carrier jitter (CCJ) signal, and a signal to noise ratio (SNR) signal.

16. A method as set forth in claim 14, wherein said audio signals are stereo signals.

17. A method as set forth in claim 15, wherein the audio signals are divided into a number of bands, each band being the source of signals from which said parameters are determined.

18. Apparatus for processing electromagnetic signals representing program and commercial signals comprising:

means adapted to be coupled to a source of said signals for receiving the signals as a signal stream;

means coupled with the receiving means for forming the signal stream into segments which can be represented as parameters, with the groups being mixed together;

means coupled with the forming means for storing the signal parameters after they have been formed;

a neural network coupled with said storing means and including classifying means operable to generate at least one decisional index representative of at least one signal segment and its parameters of at least one of said groups of signals;

means defining a transition indicator for coupling the storage means to the classifying means, said transition indicator being operable for actuating the neural network to cause a decision to be made by the neural network as to the classification of at least one signal segment, relative to said program and commercial signals;

the video signals being divided into a number of bands, each band being the source of signals from which characteristic numbers are determined; and

a signal controller coupled with the neural network for receiving and being responsive to a decisional index from the neural network, said controller being operable to control a signal sensitive structure adapted for directing said groups of signals to respective end use locations.

19. Apparatus for processing electromagnetic signals representing program and commercial signals comprising:

means adapted to be coupled to a source of said signals for receiving the signals as a signal stream;

means coupled with the receiving means for forming the signal stream into segments which can be represented as parameters, with the groups being mixed together;

means coupled with the forming means for storing the signal parameters after they have been formed;

a neural network coupled with said storing means and including classifying means operable to generate at least one decisional index representative of at least one signal segment and its parameters of at least one of said groups of signals;

means defining a transition indicator for coupling the storage means to the classifying means, said transition indicator being operable for actuating the neural network to cause a decision to be made by the neural network as to the classification of at least one signal segment, relative to said program and commercial signals;

the chrominance and luminance signals being derived from the video signals and divided into a number of spatial bands from which several characteristic numbers can be determined; and

a signal controller coupled with the neural network for receiving and being responsive to a decisional index from the neural network, said controller being operable to control a signal sensitive structure adapted for directing said groups of signals to respective end use locations.

20. A method for processing electromagnetic signals representing program and commercial television signals comprising:

receiving the signals as a signal stream at a first location;

forming the signal stream into segments which can be represented as parameters, with the groups being mixed together;

storing the signal parameters after they have been formed;

operating a neural network after the storing step to classify at least one signal segment and to generate at least one decisional index representative of at least one signal segment and their characteristic of at least one of said groups of signals;

sensing a transition from one group to the other group by actuating the neural network to cause a decision to be made by the neural network as to the classification of at least one signal segment, relative to said program and commercial signals;

storing past decisions;

allowing such past decisions to be included in the classifying decisions to determine which of said signal groups the signal segments belong to, the video signals being divided into a number of bands, each band being the source of signals from which its parameters are determined;

the chrominance and luminance signals being derived from the video signals and divided into a number of spatial bands from which several characteristic numbers can be determined; and

controlling a signal-sensitive structure for directing said groups of signals to respective end use locations.

21. A method for processing electromagnetic signals representing program and commercial television signals comprising:

receiving the signals as a signal stream at a first location;

forming the signal stream into segments which can be represented as parameters, with the groups being mixed together;

storing the signal parameters after they have been formed;

operating a neural network after the storing step to classify at least one signal segment and to generate at least one decisional index representative of at least one signal segment and their characteristic of at least one of said groups of signals;

sensing a transition from one group to the other group by actuating the neural network to cause a decision to be made by the neural network as to the classification of at least one signal segment, relative to said program and commercial signals;

storing past decisions;

allowing such past decisions to be included in the classifying decisions to determine which of said signal groups the signal segments belong to; and

controlling a signal-sensitive structure for directing said groups of signals to respective end use locations; and

chrominance and luminance signals being derived from the television signals and divided into a number of spatial bands from which several parameters can be determined.