Method of grouping images from a video sequence

1. Method of clustering images of a video sequence consisting of shots and represented by a graph-like structure, a node of the graph representing a shot or a class of shots defined by key images and the nodes being connected by edges, comprising the following iteration:

selecting an edge a k connecting nodes n i and n j

calculating a potential of node n m ,

merging of the two nodes n i and n j , as a function of the distances between the attributes of the key images defining the class of shots of node n i and those of the key images defining the class of shots of node n j and as a function of the temporal distance of these key images,

calculating a potential of each edge connecting the merged node to another node of the graph previously connected to nodes n i or n j , as a function of the distances between the attributes of the key images defining the class of shots of the merged node and those of the key images defining the class of shots of the other node and as a function of the temporal distance between these key images, the new class of shots associated with the merged node comprising the key images of the classes of shots of the merged nodes, and

merging of the two nodes and validation of the new graph if the energy of this graph, which is the sum of the potentials of the nodes and of the edges, is less than the energy of the graph before merging.

2. Method according to claim 1 , wherein the graph is initialized by assigning a node to each shot and in that edges are created from one node to another node if the shots relating to these nodes are separated by a predetermined maximum number T of shots.

3. Method according to claim 1 , wherein the value of potential associated with the edges and with the nodes is a respectively decreasing and increasing function of temporal distance.

4. Method according to claim 2 , wherein the potential of an edge connecting two nodes n i and n j is calculated according to the following equation:

V

a

⁡

(

n

i

,

n

j

)

=

min

I

k

∈

P

i

,

I

l

∈

P

j

⁢

(

w

att

·

F

a

⁡

(

K

att

,

d

~

att

,

d

att

⁡

(

I

k

,

I

l

)

)

+

w

T

·

F

a

⁡

(

K

T

,

d

~

T

,

d

T

⁡

(

I

k

,

I

l

)

)

)

and in that the potential of a node born from the merging of two nodes ni and nj is defined by the following function:

V

n

⁡

(

n

m

=

n

i

⋃

n

j

)

=

min

I

k

∈

P

i

,

I

l

∈

P

j

⁢

(

w

att

·

F

n

⁡

(

K

att

,

d

~

att

,

d

att

⁡

(

I

k

,

I

l

)

)

+

w

T

·

F

n

⁡

(

K

T

,

d

~

T

,

d

T

⁡

(

I

k

,

I

l

)

)

)

where:

P i represents the class of shots associated with node n i and comprising images I k

F a is a decreasing function

F n is an increasing function

w att and w T represent the weights related to attributes and to time

d att (I k ,I l ) is the distance function associated with the characteristic attributes of the two images

{tilde over (d)} att represents the mean of the distances between images calculated on the initial graph

d T (I k ,I l ) is the temporal distance separating the two images

{tilde over (d)} T is equal to T which represents the temporal threshold

K att and K T are two constants.

5. Method according to claim 4 , wherein the decreasing function is of the form

F

a

⁡

(

K

,

d

~

,

d

)

=

1

1

+

ⅇ

K

d

~

⁢

(

d

-

d

~

)

and in that the increasing function is of the form

F

n

⁡

(

K

,

d

~

,

d

)

=

1

-

1

1

+

ⅇ

K

d

~

⁢

(

d

-

d

~

)

.

6. Method according to claim 1 , wherein the iterations are stopped as soon as the potential merging of two nodes gives rise to an increase in the energy, the edges being selected, for the calculation of the mergings, in decreasing order of their value of potential.