Digital video processing

• US 6,026,179 A
• Filed: 07/01/1996
• Issued: 02/15/2000
• Est. Priority Date: 10/28/1993
• Status: Expired due to Fees

• Current Assignee: Media Concepts Limited
• Original Assignee: Pandora International Limited
• Inventors: Brett, Steven
• Primary Examiner(s): Au, Amelia
• Assistant Examiner(s): AHMED, SAMIR ANWAR
• Application Number:
  US08/640,886
• Time in Patent Office: 1,324 Days
• Field of Search: 382/162, 382/164, 382/165, 382/166, 382/167, 382/236, 382/274, 382/275, 382/170, 348/577, 348/579, 348/576, 348/578, 348/588, 348/587, 348/892, 358/518, 358/520, 358/515
• US Class Current: 382/162
• CPC Class Codes: H04N 9/11 Scanning of colour motion picture films, e.g. for telecine
• CPC Class Codes: H04N 9/643 Hue control means, e.g. flesh tone control
• CPC Class Codes: H04N 9/74 for obtaining special effects H04N9/65 - H04N9/73 take precedence

Abstract

A method of digitally processing a sequence of video frames wherein in object which appears in the frames and undergoes relative motion or transformation is selected in a first frame by an operator; the pixels relating to the object are tagged in that frame by means of information including at least one color or appearance attribute; corresponding pixels relating to the object are located automatically in subsequent frames, by means of said information including at least one color or appearance attribute and by means of information indicating the expected position or shape of the object in the subsequent frames; and the pixels relating to the object in each of the frames are procesed.

Claims

1. A method of digitally processing a sequence of video frames, the sequence of video frames including a representation of an object which appears in said sequence of frames and undergoes at least one of relative motion and transformation, the object being represented in each of said frames by a plurality of pixels, said method comprising the steps of:

selecting said object in a first frame under the control of an operator;

locating, and tagging the pixels defining said object in said first frame by means of information including at least one of a colour attribute and an appearance attribute;

automatically locating corresponding pixels defining said object in subsequent frames of said sequence by means of said information including at least one of a colour attribute and an appearance attribute and by means of information indicating at least one of an expected position and an expected shape of said object in said subsequent frames; and

processing only the pixels defining said object in each of said frames and passing the unselected pixels outside the object without any processing.

2. The method of claim 1, wherein the step of selecting said object in said first frame comprises the steps of:

marking a plurality of points around said object;

defining vectors joining said points so as to define a boundary around said object;

carrying out a scanning operation so as to identify pixels within said boundary; and

storing the locations of pixels identified by said scanning operation.

3. The method of claim 1 wherein said information indicating the expected position of said object is obtained by marking said object in two spaced apart frames of said sequence and interpolating to establish the expected position of said object in frames intermediate said two spaced apart frame.

4. The method of claim 1, 2 or 3 wherein the processing step comprises colour correction.

5. The method of claim 1, 2 or 3 wherein the processing step comprises spatial manipulation of said object.

6. A method of digitally processing a sequence of video frames, each of said frames comprising a plurality of digital pixels, a number of which define an object that undergoes at least one of relative motion and transformation in the course of said sequence, the method comprising the steps of:

receiving an input from an operator, said input indicating a region of a first flame containing said object;

locating and tagging, by means of information including at least one of a colour attribute and an appearance attribute of said pixels, those pixels within said region that define said object;

automatically locating corresponding pixels defining said object in subsequent frames of said sequence by means of said information including at least one of a colour attribute and an appearance attribute and by means of information indicating at least one of an expected position and an expected shape of said object in said subsequent flames; and

processing only said pixels defining said object in each of said frames and passing the unselected pixels outside the object without any processing.