Method and apparatus for detecting dissolve regions in digital video sequences

Claim 3

The method of claim 1, wherein step (c) further comprises the step of: selecting at least one pair of consecutive detected boundaries and for said at least one pair of consecutive detected boundaries i. calculating the average value of said dissolve measure between said consecutive detected boundaries; and ii. flagging a dissolve region when said average value exceeds a second predetermined threshold.

Claim 4

The method of claim 2, wherein a dissolve region is flagged in step (c)(ii) only when the distance between said consecutive detected boundaries is less than a third predetermined threshold.

Claim 5

The method of claim 3, wherein the width of said detected dissolve region in step (c)(ii) spans between and including said detected consecutive boundaries.

Claim 6

The method of claim 3, wherein the width of said dissolve region in step (c)(ii) is greater than the width spanning between said detected consecutive boundaries.

Claim 7

The method of claim 4, wherein dissolve region boundaries are detected in step (c) at spikes where the absolute value of said dissolve measure exceeds said first predetermined threshold.

Claim 8

The method of claim 6, wherein step (b) further comprises removing from said dissolve measure all but the first spike in any cluster of spikes.

Claim 9

The method of claim 7, wherein step (b) further comprises setting the value of said dissolve measure to zero at each point immediately next to a negative spike.

Claim 10

The method of claim 1, wherein said dissolve measure is calculated in step (b) based on the statistical variance of pixel values within each of said selected frames.

Claim 11

The method of claim 9, wherein said dissolve measure is the second order difference of the statistical variance of pixel values within each of said selected frames.

Claim 12

The method of claim 10, wherein said first predetermined threshold varies in accordance with the local characteristics of said dissolve measure and step (c) further comprises determining said first predetermined threshold for at least one point of said dissolve measure.

Claim 13

The method of claim 11, wherein step (c) further comprises the steps of: i. calculating a first mean representing the mean of said dissolve measure within a first window adjacent to said at least one point; ii. calculating a second mean representing the mean of said dissolve measure within a second window adjacent to said at least one point; iii. calculating the difference of the result of step (i) and the result of step (ii); and iv. comparing said difference to a fourth predetermined threshold and if said difference exceeds said fourth predetermined threshold then determining said first predetermined threshold from the mean and standard deviation of said dissolve measure within a window centered about said at least one point but excluding said at least on point otherwise determining said first predetermined threshold from the mean and standard deviation of said dissolve measure within the adjacent window corresponding to the smaller of said first and second means.

Claim 14

The method of claim 6, wherein step (b) further comprises the steps of: i. detecting all spikes in said dissolve measure which are due to hard scene changes; and ii. removing from said dissolve measure all spikes detected in step (i).

Claim 15

The method of claim 10, wherein said detected boundaries have been frame shifted and step (c) further comprises adjusting said detected boundaries to compensate for said frame shifting.

Claim 16

The method of claim 1, wherein said statistical data is derived from motion compensation displacement vectors generated from said selected frames.

Claim 17

The method of claim 1, wherein said statistical data is derived from difference images generated from said selected frames.

Claim 18

The method of claim 1, wherein said statistical data is derived from motion compensation error images generated from said selected frames.

Claim 19

The method of claim 1, wherein said statistical data is derived from the high frequency content of said selected frames.

Claim 20

The method of claim 1, wherein said statistical data is derived from the low frequency content of said selected frames.

Claim 23

The method of claim 20, wherein step (c) further comprises the step of: selecting at least one pair of consecutive detected boundaries and for said at least one pair of consecutive detected boundaries i. determining the distance between said consecutive detected boundaries; and ii. flagging a dissolve region when said distance between said consecutive detected boundaries is less than a predetermined threshold.

Claim 24

The method of claim 21, wherein the width of said dissolve region in step (c)(ii) is equal to the width between and including said detected consecutive boundaries.

Claim 25

The method of claim 21, wherein the width of said dissolve region in step (c)(ii) is greater than the width spanning between and including said detected consecutive boundaries.

Claim 26

The method of claim 20, wherein step (b) further comprises the step of filtering said dissolve measure.

Claim 27

The method of claim 20, wherein said dissolve measure is the statistical variance of pixel values with each of said selected frames.

Claim 28

The method of claim 20, wherein said statistical data is derived from motion compensation displacement vectors generated from said selected frames.

Claim 29

The method of claim 20, wherein said statistical data is derived from difference images generated from said selected frames.

Claim 30

The method of claim 20, wherein said statistical data is derived from motion compensation error images generated from said selected frames.

Claim 31

The method of claim 20, wherein said statistical data is derived from the high frequency content of said selected frames.

Claim 32

The method of claim 20, wherein said statistical data is derived from the low frequency content of said selected frames.

Claim 35

The apparatus of claim 31, wherein said dissolve region boundary detection means detects a dissolve region boundary where said dissolve measure has an absolute value which exceeds a first predetermined threshold.

Claim 36

The apparatus of claim 32, wherein said dissolve region boundary detection means further comprises: (i.) means for selecting at least one pair of consecutive detected boundaries; (ii.) means for calculating the average value of said dissolve measure between said at least one pair of consecutive detected boundaries; and (iii.) dissolve region flagging means for determining when said average value exceeds a second predetermined threshold.

Claim 37

The apparatus of claim 31, wherein said dissolve region boundary detection means detects a dissolve region boundary where said dissolve measure first increases then decreases in a parabolic fashion.

Claim 38

The apparatus of claim 34, wherein said dissolve region boundary detection means further comprises: (i.) means for selecting at least one pair of consecutive detected boundaries; (ii.) means for calculating eh average value of said dissolve measure between said at least one pair of consecutive detected boundaries; and (iii.) dissolve region flagging means for determining when said average value exceeds a second predetermined threshold.