Sub-nyquist sampling encoder and decoder of a video system

• US 4,868,654 A
• Filed: 02/29/1988
• Issued: 09/19/1989
• Est. Priority Date: 03/03/1987
• Status: Expired due to Term

• Current Assignee: Matsushita Electric Industrial Company Limited (Panasonic Holdings Corporation)
• Original Assignee: Matsushita Electric Industrial Company Limited (Panasonic Holdings Corporation)
• Inventors: Juri, Tatsuro, Etoh, Minoru
• Primary Examiner(s): Groody, James J.
• Assistant Examiner(s): Kostak, Victor R.
• Application Number:
  US07/161,928
• Time in Patent Office: 568 Days
• Field of Search: 358/138, 358/133, 358/135, 358/141, 358/12, 358/13, 340/728
• US Class Current: 375/240.21
• CPC Class Codes: H04N 19/587 involving temporal sub-sampling or interpolation, e.g. decimation or subsequent interpolation of pictures in a video sequence

Abstract

A sub-Nyguist sampling encoder and decoder has a plurality of interpolators provided in an encoder and corresponding plural interpolators provided in a decoder. Information indicating which interpolator can be used to minimize the interpolation error is superimposed on a value of a non-thinned-out pixel and transmitted from the encoder to the decoder, and the decoder receiving the information is permitted to always perform optimum interpolation. The interpolation information is superimposed on the least significant bit of the value of the non-thinned-out pixel, with the view of effecting the superimposition without increasing the amount of transmission data and without degrading quality of the video signal.

Claims

1. A sub-Nyquist sampling encoder comprising:

means for thinning out pixels from a sampled video data signal in accordance with a predetermined rule;

a plurality of interpolation means for interpolating a pixel thinned out by said thin-out means by using non-thinned-out pixels adjacent to the thinned-out pixel;

selection means for comparing a sampled value of said thinned-out pixel with a plurality of interpolation values produce from said plurality of interpolation means and selecting one of said plurality of interpolation means;

coding means responsive to an output signal of said selection means to produce a coded word indicating which interpolation means is selected by said selection means; and

superimposition means for superimposing said coded word on a non-thinned-out pixel.

2. A sub-Nyquist sampling encoder according to claim 1 wherein said plurality of interpolation means comprise a horizontal low-pass filter and a vertical low-pass filter.

3. A sub-Nyquist sampling encoder according to claim 1 wherein said superimposition means changes the value of the pixel to be superimposed in such a manner that the interpolation error is not increased by superimposing said coded word on said pixel.

4. A sub-Nyquist sampling encoder according to claim 1 wherein said superimposition means does not effect the superimposition of said coded word when the value of the pixel to be interpolated falls within the filter pass band common to all of said interpolation means.

5. A sub-Nyquist sampling encoder according to claim 4 wherein said superimposition means does not effect the superimposition of said coded word when the difference between the plurality of interpolation values is less than a predetermined value.

6. A sub-Nyquist sampling encoder according to claim 1 wherein said superimposition means superimposes said coded word on the least significant bit of a non-thinned-out pixel.

7. A sub-Nyquist sampling encoder according to claim 6 wherein said superimposition means effects the superimposition in such a manner that said coded word is represented by information as to whether the least significant bits of values of two non-thinned-out pixels are equal to each other.

8. A sub-Nyquist sampling encoder according to claim 1 wherein said superimposition means superimposes said coded word on a non-thinned-out pixel adjacent to the interpolated pixel.

9. A sub-Nyquist sampling encoder according to claim 1 further comprising pre-filtering means for cutting off the skewed frequency component of the video signal in advance of the extraction of interpolation values.

10. A sub-Nyquist sampling encoder according to claim 9 wherein said pre-filtering means comprises a skewed frequency component pass filter for producing an output signal representative of the product of horizontal high-pass filtering and vertical high-pass filtering and a subtractor for subtracting said product signal from the original video signal so as to cut off said skewed frequency component.

11. A sub-Nyquist sampling encoder as in claim 1, wherein said coding means comprises means for producing a code word having no redundant information and indicating which interpolating means is selected by said selection means.

15. A sub-Nyquist sampling encoder and decoder as in claim 11, wherein said superimposition means superimposes said coded word on the least significant bit of a non-thinned-out pixel in such a manner said coded word is represented by information as to whether the least significant bits of values of two non-thinned-out pixels are equal to each other.

12. A sub-Nyquist sampling encoder and decoder comprising:

an encoder including means for thinning out pixels from an sampled video data signal in accordance with a predetermined rule, a plurality of first interpolation means for interpolating a pixel thinned out by said thin-out means by using non-thinned-out pixels adjacent to the thinned-out pixel, selection means for comparing a sampled value of said thinned-out pixel with a plurality of interpolation values produced form said plurality of first interpolation means and selecting one of said plurality of first interpolation means, coding means responsive to an output signal of said selection means to product a coded word indicating which first interpolation means is selected by said selection means, and superimposition means for superimposing said coded word on a non-thinned-out pixel; and

a decoder, responsive to information transmitted from said encoder, including a plurality of second interpolation means having the same respective characteristics as those of corresponding ones of said first interpolation means of said encoder, separation means for separating said coded word from the non-thinned-out pixel, and insertion means for responding to said coded word to select one of said plurality of second interpolation means and inserting an interpolation value obtained from the selected second interpolation means into a position between non-thinned-out pixels.

13. A sub-Nyquist sampling encoder and decoder according to claim 12 wherein said superimposition means does not effect the superimposition of said coded word when the value of the pixel to be interpolated falls within the filter pass band common to all of said first interpolation means.

14. A sub-Nyquist sampling encoder and decoder according to claim 13 wherein said superimposition means does not effect the superimposition of said coded word when the difference between the plurality of interpolation values is less than a predetermined value.

16. A sub-Nyquist sampling encoder and decoder according to claim 12 further comprising pre-filtering means for cutting off the skewed frequency component of the video signal in advance of the extraction of interpolation values.

17. A sub-Nyquist sampling encoder and decoder according to claim 16 wherein said pre-filtering means comprises a skewed frequency component pass filter for producing an output signal representative of the product of horizontal high-pass filtering and vertical high-pass filtering and a subtractor for subtracting said product signal from the original video signal so as to cut off said skewed frequency component.