摘要 |
The invention relates to a method of processing images to determine at least one signature of a film, characterized in that it includes a step of selecting chronologically a series of images from said film, a step of evaluating shifts relating to at least one descriptor corresponding to a physical property between consecutive images in order to obtain primary visual shifts, a step of calculating differences between said primary visual shifts to determine secondary visual shifts, a step of calculating differences between said secondary visual shifts to determine tertiary visual shifts, and a step of determining a mean value and a standard deviation of the primary, secondary, and tertiary visual shifts to obtain the signature of the film. |