Sub-band coding of images with low computational complexity

• US 4,829,378 A
• Filed: 06/09/1988
• Issued: 05/09/1989
• Est. Priority Date: 06/09/1988
• Status: Expired due to Term

• Current Assignee: Telcordia Technologies Incorporated (Telefonaktiebolaget LM Ericsson)
• Original Assignee: Bell Communications Research, Inc. (Telefonaktiebolaget LM Ericsson)
• Inventors: LeGall, Didier J.
• Primary Examiner(s): Coles, Sr., Edward L.
• Assistant Examiner(s): Parker, Michael D.
• Application Number:
  US07/204,626
• Time in Patent Office: 334 Days
• Field of Search: 358/12, 358/13, 358/23, 358/30, 358/36, 358/37, 358/133, 358/136, 358/138, 358/141, 358/142, 358/143, 358/144, 358/145, 358/160, 358/166, 358/167, 358/280, 358/260, 358/261 , 358/262, 364/724.13, 364/724.16, 364/724.17, 382/41, 382/43
• US Class Current: 375/240.11
• CPC Class Codes: H03H 17/0266 Filter banks
• CPC Class Codes: H04N 19/63 using sub-band based transform, e.g. wavelets

Abstract

A class of filters for use in connection with the sub-band coding of a video signal is disclosed. The filters are of short length, symmetric and have coefficients in the form of an integer divided by a power of two. The filters may be implemented with a minimum of circuitry and permit exact reconstruction of a sub-band coded video signal.

Claims

1. A method for coding video signals comprising the steps of

dividing said video signal into a plurality of sub-band signals by applying said video signal to a first plurality of filters each of which is a symmetric finite impulse response filter comprising between three and eight stages and having coefficients which are in the form of an integer divided by a power of two, and

reconstructing said video signal by applying said sub-band signals to a second plurality of filters each of which is a symmetric finite impulse response filter comprising between three and eight stages and having coefficients which are of the form of an integer divided by a power of two.

2. The method fo claim 1 wherein said sub-band signals are decimated after processing by said first plurality of filters.

3. The method of claim 2 wherein said sub-band signals are interpolated before being processed by said second plurality of filters.

4. The method of claim 1 wherein said first plurality of filters comprises at least one pair of filters having transfer functions H₀ (z)H₁ (z) and said second plurality of filters comprises at least one pair of filters having transfer functions G₀ (z)=H₁ (-z) and G₁ (z)=-H₀ (-z) whereby said second plurality of filters serves to produce a substantially exact reconstruction of said video signal from said sub-band signals.

5. The method of claim 1 wherein said first plurality of filters comprises three pairs of filters to divide said video signal into four sub-bands and said second plurality of filters comprises three pairs of filters.

6. A device for coding a video signal comprising

at least one bank of filters for receiving said video signal and for analyzing said video signal into a plurality of sub-band signals, each of said filters being a symmetric finite impulse response filter having between three and eight stages and having coefficients in the form of an integer divided by a power of two,

means for decimating said sub-band signals, and

means for separatley coding each of said sub-band signals.

7. The device of claim 6 wherein said device further comprises a second bank of filters for further analyzing said sub-band signals, each of said filters in said second bank being a finite impulse response filter having between three and eight stages and having coefficients in the form of an integer divided by a power of two.

8. A sub-band coding system for a video signal comprising

a first filter bank for analyzing said video signal into a plurality of sub-band signals, each of said filters in said first bank being a symmetric finite impulse response filter having between three and eight stages and having coefficients in the form of an integer divided by a power of two, and

a second filter bank for synthesizing said video signal from said sub-band signals, each of said filters in said second bank being a symmetric finite impulse response filter having between three and eight stages and having coefficients in the form of an integer divided by a power of two.