Format conversion preprocessing method and circuit

• US 5,049,993 A
• Filed: 10/03/1990
• Issued: 09/17/1991
• Est. Priority Date: 10/03/1990
• Status: Expired due to Term

• Current Assignee: TTI Inventions C LLC (Intellectual Ventures LLC)
• Original Assignee: Bell Communications Research, Inc. (Telefonaktiebolaget LM Ericsson)
• Inventors: Jeng, Fure-Ching, LeGall, Didier J.
• Primary Examiner(s): Britton, Howard W.
• Application Number:
  US07/592,195
• Time in Patent Office: 349 Days
• Field of Search: 358/140, 358/138, 358/11, 358/12
• US Class Current: 348/448
• CPC Class Codes: H04N 19/587 involving temporal sub-sampling or interpolation, e.g. decimation or subsequent interpolation of pictures in a video sequence
• CPC Class Codes: H04N 19/80 Details of filtering operations specially adapted for video compression, e.g. for pixel interpolation H04N19/635, H04N19/86 take precedence
• CPC Class Codes: H04N 7/015 High-definition television systems

Abstract

A preprocessing method and circuit is utilized to convert a video sequence between a high resolution interlaced format used when the video sequence is read out of a storage medium and a low resolution progressive format suitable for compression encoding at a low bit rate. In accordance with the inventive format conversion preprocessing algorithm, the video sequence in the high resolution interlaced format is first processed by a linear filter with a diamond-shaped passband in y-t space. The output of the linear filter is processed by a non-linear filter and the resulting signal is then decimated to produce the video signal in the low resolution progressive format. The inventive preprocessing algorithm serves to eliminate aliasing, noise, and jerkiness of motion from the sequence in the low resolution progressive format.

Claims

1. A method for converting between a high resolution interlaced video sequence having two fields per frame and a low resolution progressive video sequence comprising the steps of

filtering said interlaced sequence in the vertical-time domain with a low-pass linear filter having a diamond-shaped pass-band to form a first intermediate progressive sequence,

filtering said first intermediate progressive sequence with a non-linear filter to form a second intermediate progressive sequence, and

decimating said second intermediate progressive sequence to form said low resolution progressive sequence.

2. The method of claim 1 wherein each sample s(m,n,t) in a frame of said first intermediate progressive sequence is formed by said linear filter as a weighted average of samples in a plurality of successive fields of said high resolution interlaced video sequence.

3. The method of claim 1 wherein each sample s(m,n,t) in a frame of said first intermediate progressive sequence is formed by applying said linear filter to each sample s(m,n,t) in one field of a frame of said high resolution interlaced video sequence and forming a weighted average of s(m,n,t) and adjacent samples in the preceding and succeeding fields of the interlaced sequence.

4. The method of claim 3 wherein each sample p(m,n,t) in a frame of said second intermediate progressive sequence is equal to a corresponding sample s(m,n,t) from a frame of the first intermediate sequence, or equal to a corresponding sample s(m,n,t) from a field of the high resolution interlaced sequence, or equal to a linear combination of the corresponding samples from the first intermediate progressive sequence and the high resolution interlaced sequence, depending on the absolute value of the difference between the sample from the first intermediate sequence and the sample from the high resolution interlaced sequence.

5. A circuit for converting between a high resolution interlaced video sequence and a low resolution progressive video sequence comprising,

linear filter means having a diamond-shaped pass band for filtering said interlaced sequence in the vertical-time domain, to form a first intermediate progressive sequence,

a non-linear filter for filtering said first intermediate progressive sequence to form a second intermediate progressive sequence, and

means for decimating said second intermediate progressive sequence to form said low resolution progressive sequence.

6. The circuit of claim 5 wherein said linear filter forms a sample in a frame of said first intermediate sequence by forming a weighted average of a sample in one field of a frame of said interlaced sequence and neighboring samples in the preceding and succeeding fields of the interlaced sequence.

7. The circuit of claim 6 wherein the value of an output sample of said non-linear filter in said second intermediate sequence is determined by the absolute value of the difference between a sample in said first intermediate sequence and a corresponding sample in a field of said interlaced sequence.