Digital image sensor, image capture and reconstruction method and system for implementing same

1. A digital image acquisition sensor ( 18 ) comprising an array ( 16 ) of colour filters, the array consisting of a plurality of identical basic patterns ( 70 ) which are replicated with no overlap along a vertical direction and a horizontal direction, each basic pattern being formed by colour filters ( 72 , 82 , 84 ), said color filters being pseudo-randomly arranged inside said basic pattern, such that each basic pattern ( 70 ) contains a variable pitch between two consecutive same-type colour filters in the horizontal and/or vertical directions of the basic pattern.

2. A sensor according to claim 1 , wherein the basic pattern ( 70 ) of the colour filters of the array of colour filters is equal to or larger than 4×4 pixels of an image and is smaller than the size of said image.

3. A method for acquiring and reconstructing digital images from a sensor comprising an array ( 16 ) of colour filters, the array consisting of a plurality of identical basic patterns ( 70 ) which are replicated with no overlap, each basic pattern being formed by colour filters ( 72 , 82 , 84 ) which are pseudo-randomly arranged inside said basic pattern such that each basic pattern ( 70 ) contains a variable pitch between two consecutive same-type colour filters in the horizontal and/or vertical directions of the basic pattern, said method comprising filtering said images ( 14 ) during the acquisition thereof by said sensor ( 18 ) by means of said array of colour filters ( 16 ), thus carrying out acquisition of the red, green and blue components for reconstruction of at least one image ( 30 ), from said components acquired from each image ( 22 ).

4. A method according to claim 3 , wherein a luminance and chrominance decomposition of the acquired image ( 22 ) is carried out.

5. A method according to claim 3 , wherein a luminance and chrominance decomposition of the acquired image ( 22 ) is carried out, by means of the sensor and by means of a second sensor free of an array of colour filters for acquisition of the luminance.

6. A method according to any one of claim 4 or 5 , wherein the luminance is estimated and the three components of the chrominance are interpolated from the data acquired by the pseudo-random pattern sensor ( 18 ).

7. A method according to claim 4 , wherein the reconstruction of the image ( 30 ) is carried out from the estimated luminance and the three components of the chrominance interpolated from the acquired image ( 22 ).

8. A method according claim 4 , wherein the reconstruction of the image ( 30 ) is linear and only involves a weighting of a neighbourhood of pixels ( 72 ), ( 82 ) and ( 84 ), from the basic pattern ( 70 ) and from a neighbourhood ( 76 ).

9. A method according to claim 4 , wherein the reconstruction of the image ( 30 ) is adaptive and non-linear, and carries out an interpolation of the chrominances of the spectrum ( 88 ) brought back into the baseband, via a calculation based on the gradient of the luminance ( 86 ) estimated by a standardised low-pass filter, said gradient of the low frequencies of the luminance ( 86 ) indicating the vertical or horizontal direction of interpolation of the three chrominances.

10. A method according to claim 9 , wherein demosaicing is carried out by eliminating high-risk frequencies by transfer of carriers of said zero spatial frequency chrominances.

11. An image acquisition system for implementing the method according to claim 4 , comprising digital acquisition and processing modules connected with one another, including the sensor ( 18 ), a compression module ( 32 ) connected to a reconstruction module ( 26 ), a storage module ( 40 ), a display module ( 44 ) and an output interface ( 52 ).

12. A digital image acquisition and reconstruction system according to claim 11 for implementing the method wherein the reconstruction of the image ( 30 ) is linear and only involves a weighting of a neighbourhood of pixels ( 72 ), ( 82 ) and ( 84 ), from the basic pattern ( 70 ) and from a neighbourhood ( 76 ), wherein the image ( 30 ) is reconstructed in the reconstruction module ( 26 ).

13. A digital image acquisition and reconstruction system according to claim 12 , wherein the image reconstructed ( 30 ) in the reconstruction module ( 26 ) is compressed by at least one encoder in the compression module ( 32 ).

14. A digital image acquisition and reconstruction system according to claim 12 , wherein the image reconstructed ( 30 ) in the reconstruction module ( 26 ) is compressed by at least one encoder ( 32 ) included in the reconstruction module ( 26 ).

15. A digital image acquisition and reconstruction system according to claim 11 for implementing the method, wherein the reconstruction of the image ( 30 ) is adaptive and non-linear, and carries out an interpolation of the chrominances of the spectrum ( 88 ) brought back into the baseband, via a calculation based on the gradient of the luminance ( 86 ) estimated by a standardised low-pass filter, said gradient of the low frequencies of the luminance ( 86 ) indicating the vertical or horizontal direction of interpolation of the three chrominances.

16. A digital image acquisition and reconstruction system according to claim 15 , wherein the image reconstructed ( 30 ) in the reconstruction module ( 26 ) is compressed by at least one encoder in the compression module ( 32 ).

17. A digital image acquisition and reconstruction system according to claim 15 , wherein the image reconstructed ( 30 ) in the reconstruction module ( 26 ) is compressed by at least one encoder ( 32 ) included in the reconstruction module ( 26 ).