METHOD FOR VIDEO MODE DETECTION

• US 20070263123A1
• Filed: 03/22/2007
• Published: 11/15/2007
• Est. Priority Date: 03/29/2006
• Status: Active Grant

• Current Assignee: Sony Deutschland GmbH (Sony Group Corp.)
• Original Assignee: Sony Deutschland GmbH (Sony Group Corp.)
• Inventors: Erdler, Oliver, Mayoral, Sergio
• Granted Patent: US 8,213,718 B2
• Time in Patent Office: Days
• US Class Current: 348/558
• CPC Class Codes: H04N 5/144 Movement detection for video coding H04N19/503; analysis of motion in general G06T7/20
• CPC Class Codes: H04N 5/147 Scene change detection
• CPC Class Codes: H04N 7/0115 with details on the detection of a particular field or frame pattern in the incoming video signal, e.g. 3:2 pull-down pattern
• CPC Class Codes: H04N 7/0137 dependent on presence/absence of motion, e.g. of motion zones H04N7/014 takes precedence; movement detection in television signals H04N5/144

Abstract

A method for video mode detection, wherein video input data (VID) corresponding to a video picture (P) is received and a video mode is determined for said video picture (P). The determining of said video mode depends on a local video mode (LVM) and a global video mode (GVM) of said video picture (P). Said global video mode (GVM) is determined for said video picture (P) based on said video input data (VID) or a derivative (m₁) thereof. For determining said local video mode (LVM), first said video picture (P) is subdivided into a ticker area (TA) and a remaining area (RA), thereby generating ticker area data (TAD). Then, said local video mode (LVM) is determined for said ticker area (TA) based on said ticker area data (TAD). When determining said local video mode (LVM), said ticker area (TA) is subdivided into n sub-areas, and at least one of said n sub-areas (1 . . . 6) is selected as selected sub-area (SSA). Further, a sub-area video mode (SVM) is determined for said selected sub-area and said local video mode (LVM) is determined for said ticker area (TA) based on said sub-area video mode (SVM).

Claims

1. Method for video mode detection, comprising the steps of:

determining a local video mode for a ticker area of a video input picture by

subdividing said ticker area (TA) into n sub-areas (1, . . . , 6), where n is an integer value and greater or equal to two (n≧

2),

selecting at least one of said n sub-areas (1, . . . , 6) as selected sub-area (SSA), thereby generating sub-area data (SAD),

determining a sub-area video mode (SVM) for said selected sub-area (SSA) based on said sub-area data (SAD) or a derivative (m₂) thereof,

determining said local video mode (LVM) for said ticker area (TA) based on said sub-area video mode (SVM).

2. Method according to claim 1, further comprising a step of analyzing said video picture, thereby determining a video mode of said video picture (P), wherein said video mode is determined depending on said local video mode (LVM) and a global video mode (GVM) of said video picture (P), wherein said global video mode (GVM) for said video picture (P) is determined based on said video picture (VID) or a derivative (m₁) thereof.

3. Method according to claim 1 or 2, characterized in that, when selecting at least one of said n sub-areas (1, . . . , 6), said selected sub-area (SSA) is selected depending on a previous local video mode (P-LVM), which corresponds to the local video mode (LVM) of a previous video picture.

4. Method according to any one of claims 1 to 3, characterized in that, when selecting at least one of said n sub-areas (1, . . . , 6), said selected sub-area (SSA) is selected depending on a previous sub-area video mode (P-SVM), which corresponds to the sub-area video mode of a previous video picture.

5. Method according to any one of the preceding claims, characterized in that the first sub-area (1) and the n-th sub-area (6) of said n sub-areas (1, . . . , 6,) are chosen to be smaller than the remaining n-2 sub-areas, wherein said first sub-area (1) is located on the right side of said ticker area and said n-th sub-area (6) on the left side.

6. Method according to claim 5, characterized in that in that

n is greater or equal to four,

said first sub-area (1) is completely contained in the second sub-area (2) of—

said n sub-areas (1, . . . , 6), and

said n-th sub-area (6) is completely contained in the (n-1)-th sub-area (5).

7. Method according to claim 5 or 6, characterized in that, when selecting at least one of said n sub-areas (1, . . . , 6), said first sub-area (1) and said n-th sub-area (6) are selected as selected sub-area (SSA), such that a ticker (T) running through said video picture (P) may be detected when just entering said video picture (P) from the right side or when leaving the video picture (P) on the left side.

8. Method according to any one of the preceding claims, characterized in that

the second sub-area (2) is chosen as selected sub-area (SSA) in case a previous first sub-area video mode for said first sub-area (1) was equal to camera mode,

wherein said previous first sub-area video mode corresponds to the sub-area video mode of a previous video picture within said first sub-area (1).

9. Method according to claim 8, characterized in that

if said first sub-area (1) is equal to said selected sub-area (SSA) and said sub-area video mode (SVM) is determined to be camera mode, then said local video mode (LVM) is chosen to be film mode, and subsequently,

said second sub-area (2) is chosen to be said selected sub-area (SSA) and a corresponding sub-area video mode is determined,

wherein in case said corresponding sub-area video mode is equal to camera mode, said local video mode (LVM) is chosen to be camera mode,

otherwise said local video mode (LVM) is chosen to be film mode.

10. Method according to any one of the preceding claim, characterized in that the (n-1)-th sub-area (5) is chosen as selected sub-area (SSA), in case a previous n-th sub-area video mode for said n-th sub-area (6) was camera mode, wherein said previous n-th sub-area video mode corresponds to the sub-area video mode of a previous video picture within said n-th sub-area (6).

11. Method according to claim 10, characterized in that

if the n-th sub-area (6) is equal to said selected sub-area (SSA) and said sub-area video mode (SVM) is determined to be camera mode, then said local video mode (LVM) is chosen to be film mode, and subsequently,

said (n-1)-th sub-area (5) is chosen to be said selected sub-area (SSA) and a corresponding sub-area video mode is determined,

wherein in case said corresponding sub-area video mode is equal to camera mode, said local video mode (LVM) is chosen to be camera mode,

otherwise said local video mode (LVM) is chosen to be film mode.

12. Method according to any one of the preceding claims, characterized in that

the second to (n-1)-th sub-areas (2, . . . , 5) are non-overlapping areas of said ticker area (TA), and

an m-th sub-area, where 3≦

m≦

n is selected as said selected sub-area, if a previous (m-1)-th sub-area video mode for the (m-1)-th sub-area was film mode,

wherein said previous (m-1)-th sub-area video mode corresponds to the sub-area video mode of said (m-1)-th sub-area of a previous video picture.

13. Method according to any one of the preceding claims, characterized in that

when determining said global video mode (GVM) for said video picture (P), said global video mode (GVM) is determined based on a video mode detection algorithm, and/or

when determining a sub-area video mode (SVM) for said selected sub-area (SSA), said sub-area video mode (SVM) is determined based on said video mode detection algorithm.

14. Method according to claim any one of the preceding claims, characterized in that

said derivative of said video picture corresponds to a first measurement value (m₁) indicating a degree of motion between previous and current video pictures, and

said derivative of said sub-area data corresponds to a second measurement value (m₂) indicating a degree of motion between previous and current video pictures within said sub-areas (1, . . . , 6).

15. Method according to claim 13 or 14, characterized in that said video mode detection algorithm is based on pixel differences.

16. Method according to claim 14 or 15, characterized in that said first measurement value (m₁) and/or said second measurement value (m₂) are calculated based on pixel differences between previous and current video pictures.

17. Method according to claim 13 or 14, characterized in that said video mode detection algorithm is based on a sum of motion vectors.

18. Method according to claim 14 or 17, characterized in that said first measurement value (m₁) and/or said second measurement value (m₂) are calculated based on a sum of motion vectors for previous and current video pictures.

19. Method according to any one of claims 13 to 18, characterized in that said video mode detection algorithm comprises a decision about a video mode based on a hysteresis, wherein a change of video mode is only determined after the same video mode has been determined for a predetermined number of consecutive video pictures (P).

20. Method according to any one of the preceding claims, characterized in that, when determining said local video mode (LVM) for said ticker area (TA), said local video mode (LVM) is chosen to be equal to said sub-area video mode (SVM).

21. Method according to any one of the preceding claims, characterized in that said ticker area (TA) is located in the lower, upper or middle part of said video picture (P).

22. Method according to any one of the preceding claims, characterized in that said ticker area (TA) comprises a first picture area on the lower part of said video picture and a second picture area on the upper part of said video picture.

23. Method according to any one of claims 2 to 22, characterized in that a state machine (SM) is used for modeling said analyzing of said video input picture, wherein each state of said state machine corresponds to one of said n sub-areas and to the step of determining said sub-area video mode (SVM).

24. Method according to claim 23, characterized in that a change of state within said state machine (SM) depends on the video mode determined within a previous state.

25. Method for motion compensated up-conversion depending on a video mode, wherein said video mode is determined based on a method for video mode detection as defined in claims 1 to 23.

26. Method according to claim 25, wherein a motion compensated up-conversion is performed for said ticker area (TA) depending on said local video mode (LVM), and a motion compensated up-conversion is performed for a remaining area (RA) of said video input picture depending on said global video mode (GVM), wherein said remaining area does not comprise said ticker area (TA).

27. Signal processing system,

which is capable of performing or realizing a method for video mode detection and/or a method for motion compensated up-conversion according to any one of the preceding claims 1 to 24 and/or 25 to 26 and/or the steps thereof.

28. Computer program product,

comprising computer program means adapted to perform and/or to realize a method for video mode detection and/or a method for motion compensated up-conversion according to any one of the claims 1 to 24 and/or 25 to 26 and/or the steps thereof, when it is executed on a computer, a digital signal processing means, and/or the like.

29. Computer readable storage medium,

comprising a computer program product according to claim 28.