Method of encoding and decoding the video data of an image sequence

• US 5,305,400 A
• Filed: 07/30/1992
• Issued: 04/19/1994
• Est. Priority Date: 12/05/1990
• Status: Expired due to Term

• Current Assignee: Deutsche ITT Industries GmbH
• Original Assignee: Deutsche ITT Industries GmbH
• Inventors: Butera, Bill
• Primary Examiner(s): Mancuso, Joseph
• Application Number:
  US07/931,461
• Time in Patent Office: 628 Days
• Field of Search: 382/1, 382/48, 382/36, 382/56, 358/105, 358/108, 358/133, 358/136, 358/135, 358/426
• US Class Current: 382/107
• CPC Class Codes: H04N 19/51 Motion estimation or motion compensation
• CPC Class Codes: H04N 19/90 using coding techniques not provided for in groups H04N19/10-H04N19/85, e.g. fractals
• CPC Class Codes: H04N 7/122 involving expansion and subsequent compression of a signal segment, e.g. a frame, a line

Abstract

The video data of an image sequence are encoded and decoded section by section, with each nth key frame data field being transmitted as completely as possible. From two adjacent key frame data fields, n-1 phantom image data fields are calculated at the transmitting and receiving ends taking into account n-1 motion fields. Further, n-1 difference image data fields are determined at the transmitting end and transmitted together with the key frame data fields. At the receiving end, the original sequence of data fields is reconstructed from the transmitted data fields by recalculation of the motion fields.

Claims

1. Method of encoding video data of an image sequence whose individual images follow each other at fixed time intervals, with each individual image having a frame data field associated therewith, said method comprising the steps of:

(a) dividing the sequence of frame data fields into sections each comprising n frame data fields, with a first frame data field in each section serving as a key frame data field, and the remaining n-1 frame data fields in each section serving as intermediate-frame data fields, where n is greater than or equal to 2;

(b) determining n-1 motion fields from the first key frame data field in a given section and from a second key frame data field in a next following second, by means of n-1 motion estimators, the respective reference times of said motion fields being equal to the reference times of the associated intermediate-frame data fields;

(c) computing from the first and second key frame data fields and one of the motion fields, a respective phantom image data field by means of a respective phantom-image-computing device;

(d) computing, from the phantom image data field and the associated intermediate-frame data field, difference image data by means of a subtractor; and

(e) generating an output signal as a data stream formed from the data of the key frame data fields and the difference image data, said data being inserted into the data stream by means of a transmitter switching device which also inserts at least one synchronizing character for identifying and separating the individual data fields.

2. A method of decoding video data of an image sequence, said video data being contained in an encoded data stream, the method of decoding comprising the steps of:

(a) separating, by means of a receiver switching device, a first key frame data field, a second key frame data field, and n-1 difference image data fields from said data stream including at least one synchronizing character, and feeding same to one memory device each, where n is greater than or equal to 2;

(b) a determining, by means of n-1 motion estimators, n-1 motion fields from the first and second key frame data fields, the respective reference times of said motion fields being equal to the reference times of the associated difference image data fields;

(c) a computing from one of the motion fields and the first and second key frame data fields, a respective phantom image data field by means of a respective phantom-imaged-computing device;

(d) computing, from the phantom image data field and the associated difference image data field, an associated intermediate-frame data field by means of a summing device; and

(e) forming an output signal of the first and second key frame data fields and the n-1 reconstructed intermediate-frame data fields.

3. Video data compression apparatus comprising:

(a) means for dividing a sequence of video frame data fields into sections each comprising N frame data fields, with a selected frame data field in each section serving as a key frame data field, and the remaining N-1 frame data fields in each section serving as intermediate frame data fields, where N is an integer greater than one;

(b) motion estimator means for determining N-1 motion fields from a first key frame data field in a given section and a second key frame data field in a following section, said N-1 motion fields having reference times corresponding to reference times of said intermediate frame data fields;

(c) means responsive to said first and second key frame data fields and to said N-1 motion fields for generating N-1 phantom-image data fields, each phantom-image data field corresponding to a given motion field;

(d) a subtractor responsive to said N-1 phantom-image data fields and to said N-1 intermediate frame data fields for generating N-1 difference image data fields, each difference image data field corresponding in its reference time to that of the phantom-image data field and the intermediate frame data field from which it was generated; and

(e) means for generating an output data stream formed of data corresponding to the key frame data fields and the difference image data fields.

4. The video data compression apparatus of claim 3, wherein said means for generating an output data stream inserts synchronizing data into said stream.

5. The video data compression apparatus of claim 3, wherein said motion estimator means linearly interpolates N-1 motion fields based upon the first key frame data field in said given section and the second key frame data field in said following section.

6. Video data decompression apparatus for generating frames of video data from a compressed data format, said apparatus comprising:

(a) means for separating incoming data in said compressed data format into a sequence of sections of data, each section including a key frame data field and N-1 difference image data fields, where N is an integer greater than one;

(b) motion estimator means for determining N-1 motion fields from a first key frame data field in a given section and a second key frame data field in a following section, said N-1 motion fields having reference times corresponding to reference times of said difference image data fields;

(c) means responsive to said first and second key frame data fields and to said N-1 motion fields for generating N-1 phantom-image data fields, each phantom-image data field corresponding to a given motion field;

(d) summing means responsive to said N-1 phantom-image data fields and to said N-1 difference image data fields for generating N-1 intermediate frame image data fields, each intermediate frame image data field corresponding in its reference time to that of the phantom-image data field and the difference image data field from which it was generated; and

(e) means for generating an output data stream formed of a sequence of sections of data, each section having N data fields comprising the key frame data field for that section and the N-1 intermediate image data fields for that section.

7. The video data decompression apparatus of claim 6, wherein said means for separating incoming data in said compressed data format is responsive to synchronizing data in said incoming data.

8. The video data compression apparatus of claim 6, wherein said motion estimator means linearly interpolates N-1 motion fields based upon the first key frame data field in said given section and the second key frame data field in said following section.

9. A method of compressing video data comprising the steps of:

(a) dividing a sequence of video frame data fields into sections each comprising N frame data fields, with a selected frame data field in each section serving as a key frame data field, and the remaining N-1 frame data fields in each section serving as intermediate frame data fields, where N is an integer greater than one;

(b) determining N-1 motion fields from a first key frame data field in a given section and a second key frame data field in the following section, said N-1 motion fields having reference times corresponding to reference times of said intermediate frame data fields;

(c) generating, in response to said first and second key frame data fields and to said N-1 motion fields, N-1 phantom-image data fields, each phantom-image data field corresponding to a given motion field;

(d) subtracting said N-1 phantom-image data fields and said N-1 intermediate frame data fields to generate N-1 difference image data fields, each difference image data field corresponding in its reference time to that of the phantom-image data field and the intermediate frame data field from which it was generated; and

(e) generating an output data stream formed of data corresponding to the key frame data fields and the difference image data fields.

10. The video data compression method of claim 9, wherein the step of generating an output data stream includes inserting synchronizing data into said stream.

11. The video data compression method of claim 9, wherein the step of determining N-1 motion fields linearly interpolates the N-1 motion fields based upon the first key frame data field in said given section and the second key frame data field in said following section.

12. A method for recovering video data from a compressed data format, said method comprising the steps of:

(a) separating incoming data in said compressed data format into a sequence of sections of data, each section including a key frame data field and N-1 difference image data fields, where N is an integer greater than one;

(b) determining N-1 motion fields from a first key frame data field in a given section and a second key frame data field in a following section, said N-1 motion fields having reference times corresponding to reference times of said difference image data fields;

(c) generating, in response to said first and second key frame data fields and to said N-1 motion fields, N-1 phantom-image data fields, each phantom-image data field corresponding to a given motion field;

(d) summing said N-1 phantom-image data fields and said N-1 difference image data fields to generate N-1 intermediate frame image data fields, each intermediate frame image data field corresponding in its reference time to that of the phantom-image data field and the difference image data field from which it was generated; and

(e) generating an output data stream formed of a sequence of sections of data, each section having N data fields comprising the key frame data field for that section and the N-1 intermediate image data fields for that section.

13. The video data recovery method of claim 12, wherein the step of separating incoming data in said compressed data format responds to synchronizing data in the incoming data.

14. The video data recovery method of claim 12, wherein the step of determining N-1 motion fields linearly interpolates N-1 motion fields based upon the first key frame data field in the given section and the second key frame data field in the following section.