Method for image processing and video compression with sparse zone salient features

1. A method for video compression through image processing and object detection, to be carried out by an electronic processing unit, based either on images or on a digital video stream of images, the images being defined by a single frame or by sequences of frames of said video stream, with the aim of enhancing and then isolating frequency domain signals representing a content to be identified, and decreasing or ignoring frequency domain noise with respect to the content within the images or the video stream, comprising the steps of:

obtaining a digital image or a sequence of digital images from either a corresponding single frame or a corresponding sequence of frames of said video stream, all the digital images being defined in a spatial domain;

selecting one or more pairs of sparse zones, each covering at least a portion of said single frame or at least two frames of said sequence of frames, each pair of sparse zones generating a selected feature, each zone being defined by two sequences of spatial data;

transforming the selected features into frequency domain data by combining, for each zone, said two sequences of spatial data through a 2D variation of an L-transformation, varying a respective transfer function, shape and direction of the frequency domain data for each zone, thus generating a normalized complex vector for each of said selected features;

combining all said normalized complex vectors to define a model of the content to be identified; and

inputting that model from said selected features in a classifier, therefore obtaining the data for object detection or visual saliency to use for video compression.

2. The method for video compression as defined in claim 1 , wherein the step of transforming the selected features into frequency domain data uses spatial data from a varying number and/or choice of frames.

3. The method of video compression according to claim 1 , wherein a search logic is used on the full input image to generate an input frame where said sparse zones are identified.

4. The method of video compression according to claim 1 , wherein said sparse zones are grouped together, either possibly partially overlapping each other or placed side-to-side, to increase a local resolution of said digital image at said sparse zones.

5. The method of video compression according to claim 1 , wherein the transforming the selected features into frequency domain data is carried out in parallel with respect to said two sequences of spatial data.

6. The method of video compression according to claim 1 , wherein, in the transforming step, first 1D Göertzel calculations are performed by rows and then the results are used for a second step wherein 1D Göertzel calculations are performed by columns, or vice versa.

7. The method of video compression according to claim 1 , wherein, for each sparse zone of a pair, different target frequencies are chosen.

8. The method of video compression according to claim 1 , wherein input cells of digital images for the step of transforming into the frequency domain data are only taken around a position for which a transforming computing is needed.

9. The method of video compression according to claim 8 , wherein the transforming computing of the position is taken by separately calculating the 1D output for the row and column at the position and then combining this into a single value.

10. The method of video compression according to claim 1 , wherein the transfer function is chosen separately for each input of a sparse zone, so that the first input and second input have different discrete transfer function settings.