CORRECTING DISTURBANCE IN A PIXEL SIGNAL INTRODUCED BY SIGNAL FILTERING IN A DIGITAL CAMERA
A camera is configured with a disturbance correction engine that identifies and corrects disturbance introduced in image data when a filter is applied to the image data. The disturbance may take the form of overshoot noise, crosstalk noise, and/or low pass band energy. The disturbance correction engine determines an amount of energy deficit in the pass band and compensates for the deficit using the increase in energy in the side band. In operation, the disturbance correction engine processes the filtered image data in the frequency domain to adjust the amplitudes of the image coefficients. The adjusted image coefficients compensate for the deficit of energy in the pass band and also correct the disturbance caused by the overshoot and the crosstalk.
1 . A camera, comprising:
an image sensor configured to capture light incident upon the image sensor to produce image data, the image data comprising a plurality of pixels; and
a processor configured to:
access image data associated with a pixel, the image data having a first quantity of pass band energy,
generate filtered image data using one or more filters, the filtered image data having a second quantity of pass band energy that is less than the first quantity and having excess side band energy,
transform the filtered image data to generate a set of filtered image coefficients associated with the filtered image data, the set of filtered image coefficients being a representation of the filtered image data in the frequency domain,
adjust the set of filtered image coefficients to compensate for the second quantity of energy being lower than the first quantity of energy based on the excess side band energy to produce an adjusted set of filtered image coefficients,
transform the adjusted set of filtered image coefficients into the spatial domain to produce adjusted image data, and
process the adjusted image data to produce a digital image.
2 . The camera of claim 1 , wherein the excess side band energy causes crosstalk and overshoot noise in the filtered image data, and wherein adjusting the set of filtered image coefficients comprises determining the second quantity of energy, a third quantity of energy in the crosstalk, and a fourth quantity of energy in the overshoot of the filtered image data.
3 . The camera of claim 2 , wherein adjusting the set of filtered image coefficients further comprises modifying an amplitude of at least one of the filtered image coefficients based on one or more of the second quantity of energy, the third quantity of energy, and the fourth quantity energy.
4 . The camera of claim 2 , wherein the set of filtered image coefficients includes a first order filtered image coefficient, and wherein adjusting the set of filtered image coefficients comprises computing an amount by which to adjust the first order filtered image coefficient using the formula:
f
0
uc
[
1
E
IB
-
1
]
,
where f 0uc is the amplitude of the first order filtered image coefficient and E IB is the second quantity of energy.
5 . The camera of claim 2 , wherein the set of filtered image coefficients includes a second order filtered image coefficient, and wherein adjusting the set of filtered image coefficients comprises computing an amount by which to adjust the second order filtered image coefficient using the formula
f
±
1
uc
[
1
-
(
E
O
+
E
C
)
2
]
,
where f +1uc is the amplitude of the second order filtered image coefficient, E C is the third quantity of energy, and E O is the fourth quantity of energy.
6 . The camera of claim 2 , wherein the set of filtered image coefficients includes a second order filtered image coefficient, and wherein adjusting the set of filtered image coefficients comprises computing an amount by which to adjust the second order filtered image coefficient using the formula
f
±
1
uc
[
E
IB
2
]
,
where f +1uc is the amplitude of the second order filtered image coefficient and E IB is the second quantity of energy.
7 . The camera of claim 2 , wherein the set of filtered image coefficients includes a second order filtered image coefficient, and wherein adjusting the set of filtered image coefficients comprises computing an amount by which to adjust the third order filtered image coefficient using the formula
f
±
2
uc
[
1
E
IB
·
1
2
]
,
where f +2uc is the amplitude of the third order filtered image coefficient and E IB is the second quantity of energy.
8 . The camera of claim 2 , wherein determining the second, third, and fourth quantities of energy comprises analyzing the filtered image data to measure the pass band energy, the energy in the crosstalk, and the energy in the overshoot.
9 . The camera of claim 1 , wherein determining the second, third, and fourth quantities of energy comprises identifying a type associated with each of the one or more filters, and computing the second, third, and fourth quantities of energy based on the identified types and the image data.
10 . A method for capturing an image with a camera, the method comprising:
capturing, by an image sensor of the camera, light incident upon the image sensor to produce image data, the image data comprising a plurality of pixels;
accessing image data associated with a pixel, the image data having a first quantity of pass band energy;
generating, by a filter engine, filtered image data using one or more filters, the filtered image data having a second quantity of pass band energy that is less than the first quantity and having excess side band energy;
transforming the filtered image data to generate a set of filtered image coefficients associated with the filtered image data, the set of filtered image coefficients being a representation of the filtered image data in the frequency domain;
adjusting the set of filtered image coefficients to compensate for the second quantity of energy being lower than the first quantity of energy based on the excess side band energy to produce an adjusted set of filtered image coefficients;
transforming the adjusted set of filtered image coefficients into the spatial domain to produce adjusted image data; and
processing the adjusted image data to produce a digital image.
11 . The method of claim 10 , wherein the excess side band energy causes crosstalk and overshoot noise in the filtered image data, and wherein adjusting the set of filtered image coefficients comprises determining the second quantity of energy, a third quantity of energy in the crosstalk, and a fourth quantity of energy in the overshoot of the filtered image data.
12 . The method of claim 11 , wherein adjusting the set of filtered image coefficients further comprises modifying an amplitude of at least one of the filtered image coefficients based on one or more of the second quantity of energy, the third quantity of energy, and the fourth quantity energy.
13 . The method of claim 11 , wherein the set of filtered image coefficients includes a first order filtered image coefficient, and wherein adjusting the set of filtered image coefficients comprises computing an amount by which to adjust the first order filtered image coefficient using the formula:
f
0
uc
[
1
E
IB
-
1
]
,
where f 0uc is the amplitude of the first order filtered image coefficient and E IB is the second quantity of energy.
14 . The method of claim 11 , wherein the set of filtered image coefficients includes a second order filtered image coefficient, and wherein adjusting the set of filtered image coefficients comprises computing an amount by which to adjust the second order filtered image coefficient using the formula
f
±
1
uc
[
1
-
(
E
O
+
E
C
)
2
]
,
where f +1uc is the amplitude of the second order filtered image coefficient, E C is the third quantity of energy, and E O is the fourth quantity of energy.
15 . The method of claim 11 , wherein the set of filtered image coefficients includes a second order filtered image coefficient, and wherein adjusting the set of filtered image coefficients comprises computing an amount by which to adjust the second order filtered image coefficient using the formula
f
±
1
uc
[
E
IB
2
]
,
where f +fuc is the amplitude of the second order filtered image coefficient and E IB is the second quantity of energy.
16 . The method of claim 11 , wherein the set of filtered image coefficients includes a second order filtered image coefficient, and wherein adjusting the set of filtered image coefficients comprises computing an amount by which to adjust the third order filtered image coefficient using the formula
f
±
2
uc
[
1
E
IB
·
1
2
]
,
where f +2uc is the amplitude of the third order filtered image coefficient and E IB is the second quantity of energy.
17 . The method of claim 11 , wherein determining the second, third, and fourth quantities of energy comprises analyzing the filtered image data to measure the pass band energy, the energy in the crosstalk, and the energy in the overshoot.
18 . The method of claim 10 , wherein determining the second, third, and fourth quantities of energy comprises identifying a type associated with each of the one or more filters, and computing the second, third, and fourth quantities of energy based on the identified types and the image data.
19 . A camera, comprising:
a memory for storing image data associated with a pixel, the image data having a first quantity of pass band energy;
a processor configured to:
generate filtered image data using one or more filters, the filtered image data having a second quantity of pass band energy that is less than the first quantity and having excess side band energy,
transform the filtered image data to generate a set of filtered image coefficients associated with the filtered image data, the set of filtered image coefficients being a representation of the filtered image data in the frequency domain,
adjust the set of filtered image coefficients to compensate for the second quantity of energy being lower than the first quantity of energy based on the excess side band energy to produce an adjusted set of filtered image coefficients, and
transform the adjusted set of filtered image coefficients into the spatial domain to produce adjusted image data.
20 . The camera of claim 19 , wherein the excess side band energy causes crosstalk and overshoot noise in the filtered image data, and wherein adjusting the set of filtered image coefficients comprises determining the second quantity of energy, a third quantity of energy in the crosstalk, and a fourth quantity of energy in the overshoot of the filtered image data.