Fixed pattern noise compensation with low memory requirements
View Patent ↗A method and a system for the compensation of Fixed Pattern Noise (FPN) in digital images have been achieved. The FPN compensation is based on processing done during the production of said images. The fixed pattern noise is here defined as the fixed pattern seen in the individual pixel offsets. The fixed pattern noise is uncorrelated noise but it has a statistical distribution that can be scaled to fit all images. The general idea is to measure the distribution for each individual camera, compress it, and save it in the module. For each image that is then taken with the module the noise pattern can be retrieved and rescaled to fit the image. Covered pixels are employed to normalize the FPN data to the current frame. In order to minimize memory requirements a compression scheme has to be used. A method combining a quantization step with a non-lossy compression is used. The black level is corrected for as part of the operation.
1. A method to compensate fixed pattern noise of digital images without the need of an external shutter comprising:
providing a non-volatile memory, a processing device, calibration frames, and an image sensor;
derive FPN data for each sensor using calibration frames, wherein said calibration frames are obtained at high temperatures, which are close to maximum allowable temperature of the image sensor, with long exposure time, being at the high end of the range of exposure time the image sensor is used for having the camera lens covered;
store FPN data measured;
retrieve FPN data for each picture taken; and
apply individual FPN data on each pixel of pictures taken to compensate fixed pattern noise.
2. The method of claim 1 wherein said FPN data comprise the device and interconnect parameter variations across said sensor.
3. The method of claim 1 wherein said image sensor is a CMOS image sensor.
4. The method of claim 1 wherein said image sensor is a CCD image sensor.
5. The method of claim 1 wherein said non-volatile memory is a Flash Memory.
6. The method of claim 1 wherein said non-volatile memory is an MRAM.
7. The method of claim 1 wherein said non-volatile memory is an EPROM.
8. The method of claim 1 wherein said processing device is a microprocessor.
9. A method to compensate fixed pattern noise of digital images without the need of an external shutter and with low memory requirements comprising:
providing a non-volatile memory, a processing device, calibration frames, and an image sensor;
derive FPN data for each sensor using calibration frames, wherein said calibration frames are obtained at high temperatures, which are close to maximum allowable temperature of the image sensor, with long exposure time being at the high end of the range of exposure time the image sensor is used for, having the camera lens covered;
reduce said FPN data derived;
perform compression on FPN data;
store factor used for data reduction, compressed FPN data, compression table used, and distribution measurement results in said non-volatile memory;
perform decompression on FPN data;
perform scaling of reduced FPN data using current statistical distribution measurement; and
apply individual compensation value on each pixel.
10. The method of claim 9 wherein said FPN data comprise the device and interconnect parameter variations across said sensor.
11. The method of claim 9 wherein said data reduction is performed by a quantization of said FPN data and the related quantization factor is stored together with the FPN data and said scaling of the reduced FPN data is performed using said quantization factor.
12. The method of claim 11 wherein said quantization is comprising a normalization by standard deviation σ of the calibration image and a scaling by standard deviation σ′ of the target image.
13. The method of claim 9 wherein a standard deviation is used for said distribution measurement.
14. The method of claim 9 wherein said image sensor is a CMOS image sensor.
15. The method of claim 9 wherein said image sensor is a CCD image sensor.
16. The method of claim 9 wherein said non-volatile memory is a Flash Memory.
17. The method of claim 9 wherein said non-volatile memory is an MRAM.
18. The method of claim 9 wherein said non-volatile memory is an EPROM.
19. The method of claim 9 wherein a non-lossy compression method is used for said compression and decompression of FPN data.
20. The method of claim 19 wherein said compression and decompression of FPN data is performed using the Huffman compression algorithm.
21. The method of claim 9 wherein said processing device is a microprocessor.
22. A method to compensate fixed pattern noise of digital images without the need of an external shutter and with low memory requirements comprising:
providing a non-volatile memory, a processing device, calibration frames, and an image sensor;
derive FPN data for each sensor using calibration frames, wherein said calibration frames are obtained at high temperatures, which are close to maximum allowable temperature of the image sensor, with long exposure time, being at the high end of the range of exposure time the image sensor is used for, having the camera lens covered;
quantize said FPN data derived;
generate optimized Huffman table;
perform Huffman compression on FPN data using optimized Huffman table;
store quantization factor, compressed FPN data, Huffman compression table used, and distribution measurement results;
perform Huffman decompression on FPN data;
perform quantization scaling using current statistical distribution measurement; and
apply individual compensation value on each pixel.
23. The method of claim 22 wherein pixels of said image sensor are covered with opaque material.
24. The method of claim 23 wherein said opaque material is metal.
25. The method of claim 23 wherein the impact of temperature and exposure time on the fixed pattern noise is considered by a normalization of stored FPN data by statistical data obtained from said covered pixels.
26. The method of claim 22 wherein standard deviation is used for said distribution measurement.
27. The method of claim 22 wherein said quantization scaling is comprising a normalization by standard deviation σ of the calibration image and a scaling by standard deviation σ′ of the target image.
28. The method of claim 22 wherein said image sensor is a CMOS image sensor.
29. The method of claim 22 wherein said image sensor is a CCD image sensor.
30. The method of claim 22 wherein said non-volatile memory is a Flash Memory.
31. The method of claim 22 wherein said non-volatile memory is an MRAM.
32. The method of claim 22 wherein said non-volatile memory is an EPROM.
33. The method of claim 22 wherein said processing device is a microprocessor.
34. A system to compensate fixed pattern noise of digital images without the need of an external shutter and due to FPN data compression with low memory requirements comprising:
a non-volatile memory;
a processing device being capable to reduce FPN data, to perform compression on FPN data, to perform decompression on FPN data, to perform scaling of reduced FPN data, and to apply individual compensation value on each pixel; and
an image sensor.
35. The system of claim 34 wherein said image sensor is a CMOS image sensor.
36. The system of claim 34 wherein said image sensor is a CCD image sensor.
37. The system of claim 34 wherein said non-volatile memory is a Flash Memory.
38. The system of claim 34 wherein said non-volatile memory is an MRAM.
39. The system of claim 34 wherein said non-volatile memory is an EPROM.
40. The system of claim 34 wherein said processing device is a microprocessor.
41. The system of claim 34 wherein an optional image buffer is used.
42. The system of claim 41 wherein said optional image buffer is a RAM.
43. The system of claim 34 wherein some of the pixels of said image sensor are covered with opaque material.
44. The system of claim 43 wherein said opaque material is metal.