Power reduction for image sensor with raw image scaler
A system and method for scaling an image includes receiving raw image data comprising input pixel values which correspond to pixels of an image sensor, filtering pixels in a spatial domain, and filtering pixels in a frequency domain according to an oversampling ratio. The system and method may also include outputting scaled image data as output pixel values, which correspond to subgroups of the input pixel values. The filtering may be done according to a Bayer-consistent ruleset which includes a set of filter weights and a series of scaling rules. The oversampling ratio is set to minimize an error after the filtering in the spatial domain and the filtering in the frequency domain.
1. A method of scaling an image in an image scaling circuit, comprising:
receiving, at the image scaling circuit, a raw image data comprising a plurality of input pixel values, respective ones of the input pixel values corresponding to a respective pixel of an image sensor;
filtering, by the image scaling circuit, the plurality of input pixel values in a spatial domain;
filtering, by the image scaling circuit, respective ones of the plurality of input pixel values in a frequency domain according to a corresponding predetermined oversampling ratio.
2. The method according to claim 1 , wherein the corresponding predetermined oversampling ratio is set such that an error after the filtering in the spatial domain and the filtering in the frequency domain is less than two times an error of a single sigma-delta conversion.
3. The method according to claim 1 , further comprising:
in the step of filtering in the spatial domain, converting, by the image scaling circuit, the raw image data from an analog data to a digital data.
4. The method according to claim 3 , wherein the corresponding predetermined oversampling ratio is proportional to an error in the step of converting.
5. The method according to claim 1 , further comprising:
in the step of filtering in the spatial domain, filtering, by the image scaling circuit, the plurality of input pixel values according to a Bayer-consistent ruleset.
6. The method according to claim 5 , wherein the Bayer-consistent ruleset includes:
a plurality of spatial filter weights;
a first rule that a respective output pixel value corresponds to a first color, and the corresponding subgroup of the plurality of input pixel values also corresponds to the first color;
a second rule that the plurality of input pixel values corresponds to a plurality of input sub-tiles and the spatial filter weights for the first color are concentrated within an input sub-tile of the same color; and
a third rule that a center of gravity of the spatial filter weights for the first color coincides with the geometric pixel center of the output pixel of the first color.
7. The method according to claim 1 , further comprising:
outputting, from the image scaling circuit, scaled image data comprising a plurality of output pixel values, respective ones of the output pixel values corresponding to a subgroup of the plurality of input pixel values.
8. The method according to claim 1 , wherein the corresponding predetermined oversampling ratio for a scaling factor larger than 1 is smaller than the corresponding predetermined oversampling ratio for a scaling factor equal to 1.
9. An image processing circuit, comprising:
an image scaling circuit including:
an input unit configured to receive raw image data comprising a plurality of input pixel values, respective ones of the input pixel values corresponding to a respective pixel of an image sensor; and
a filter unit configured to filter respective ones of the plurality of pixels in a spatial domain, and to filter the plurality of input pixel values in a frequency domain according to a corresponding predetermined oversampling ratio.
10. The image processing circuit according to claim 9 , wherein corresponding the predetermined oversampling ratio is set such that an error after the filtering in the spatial domain and the filtering in the frequency domain is less than two times an error of a single sigma-delta conversion.
11. The image processing circuit according to claim 9 , wherein the image scaling circuit further includes an analog-to-digital converter.
12. The image processing circuit according to claim 11 , wherein the corresponding predetermined oversampling ratio is proportional to an error in the step of converting.
13. The image processing circuit according to claim 9 , wherein the filter unit is configured to filter the plurality of pixels in the spatial domain according to a Bayer-consistent ruleset.
14. The image processing circuit according to claim 13 , wherein the Bayer-consistent ruleset includes:
a plurality of spatial filter weights;
a first rule that a respective output pixel value corresponds to a first color, and the corresponding subgroup of the plurality of input pixel values also corresponds to the first color;
a second rule that the plurality of input pixel values corresponds to a plurality of input sub-tiles and the spatial filter weights for the first color are concentrated within an input sub-tile of the same color; and
a third rule that a center of gravity of the spatial filter weights for the first color coincides with the geometric pixel center of the output pixel of the first color.
15. The image processing circuit according to claim 9 , further comprising:
an output unit configured to output scaled image data comprising a plurality of output pixel values, respective ones of the output pixel values corresponding to a subgroup of the plurality of input pixel values.
16. The image processing circuit according to claim 9 , wherein the corresponding predetermined oversampling ratio for a scaling factor larger than 1 is smaller than the corresponding predetermined oversampling ratio for a scaling factor equal to 1.
17. An imaging device, comprising:
an image sensor including a plurality of pixels; and
an image scaling circuit including:
an input unit configured to receive raw image data comprising a plurality of input pixel values, respective ones of the input pixel values corresponding to a respective pixel of the image sensor; and
a filter unit configured to filter the plurality of pixels in a spatial domain, and to filter respective ones of the plurality of input pixel values in a frequency domain according to a corresponding predetermined oversampling ratio.
18. The imaging device according to claim 17 , wherein the corresponding predetermined oversampling ratio is set such that an error after the filtering in the spatial domain and the filtering in the frequency domain is less than two times an error of a single sigma-delta conversion.
19. The imaging device according to claim 17 , wherein the image scaling circuit further includes an analog-to-digital converter.
20. The imaging device according to claim 19 , wherein the corresponding predetermined oversampling ratio is proportional to an error in the step of converting.
21. The imaging device according to claim 17 , wherein the filter unit is configured to filter the plurality of pixels in the spatial domain according to a Bayer-consistent ruleset.
22. The imaging device according to claim 21 , wherein the Bayer-consistent ruleset includes:
a plurality of spatial filter weights;
a first rule that a respective output pixel value corresponds to a first color, and the corresponding subgroup of the plurality of input pixel values also corresponds to the first color;
a second rule that the plurality of input pixel values corresponds to a plurality of input sub-tiles and the spatial filter weights for the first color are concentrated within an input sub-tile of the same color; and
a third rule that a center of gravity of the spatial filter weights for the first color coincides with the geometric pixel center of the output pixel of the first color.
23. The imaging device according to claim 17 , further comprising:
an output unit configured to output scaled image data comprising a plurality of output pixel values, respective ones of the output pixel values corresponding to a subgroup of the plurality of input pixel values.
24. The imaging device according to claim 17 , wherein the corresponding predetermined oversampling ratio for a scaling factor larger than 1 is smaller than the corresponding predetermined oversampling ratio for a scaling factor equal to 1.
25. The method according to claim 1 , wherein respective ones of the plurality of pixel values are filtered in the frequency domain by corresponding predetermined oversampling ratios different from each other.
26. The image processing circuit according to claim 9 , wherein the filter unit is configured to filter respective ones of the plurality of input pixel values in the frequency domain by corresponding predetermined oversampling ratios different from each other.
27. The imaging device according to claim 17 , wherein the filter unit is configured to filter respective ones of the plurality of input pixel values in the frequency domain by corresponding predetermined oversampling ratios different from each other.