Image sensor signal defect correction

1. A method of correcting defects in image data comprising an array of pixels, the method comprising:

sampling ( 5 ) the intensity of pixels of the image data in each side of a defect;

calculating ( 6 ) differences between the sampled intensities to generate one or more first intensity difference signals (D 1 , D 2 ) indicative of intensity differences across the defect;

calculating ( 6 ) difference between the sampled intensities to generate at least two second intensity difference signals (D 3 , D 4 ) indicative of an intensity difference on each respective side of the detect, and

correcting ( 7 - 11 ) the defect in dependence on the first and second intensity difference signals (D 1 -D 4 ), a defect being replaced by an average of surrounding pixels if said first and second intensity difference signals do not exceed a predetermined level value (L 1 ,L 2 ).

2. A method according to claim 1 , wherein the sampling step ( 5 ) comprises sampling the intensity of three pixels on each side of the defect, and wherein the calculating step ( 6 ) comprises two intensity difference signals (D 1 ,D 2 ) across a defect, and one intensity difference signal (D 3 , D 4 ) at each side of the detect.

3. A method as claimed in claim 1 , further comprising the steps of:

comparing ( 7 ) the intensity difference signals (D 1 -D 4 ) with the predetermined level value (L 1 , L 2 ) to generate ( 7 ) a first control signal (AC) if any one of the intensity difference signals (D 1 -D 4 ) is above the predetermined level value (L 1 , L 2 ), and a second control signal (DC) if all of the intensity difference signals are below the predetermined level value (L 1 , L 2 ).

4. A method according to claim 3 , wherein the predetermined level value (L 1 , L 2 ) is determined by calibration to an apparatus for which the method is used.

5. A method according to claim 3 , wherein the predetermined level signal (L 1 , L 2 ) comprises a first level signal (L 1 ) and a second level signal (L 2 ).

6. A method according to claim 3 , further comprising the additional step of assessing ( 9 ) a size of the defect if the first control signal (AC) is generated.

7. A method according to claim 6 , further comprising employing average correction techniques ( 8 ) if the first control signal (AC) is generated and the size of the defect is large (L).

8. A method according to claim 6 , further comprising employing no defect pixel value correction if the size of the defect is small (S).

9. A method according to claim 6 , further comprising the step of assessing ( 10 ) the level of good nearby pixels in the event that the defect size is medium (M).

10. A method according to claim 9 , wherein no defect pixel value correction ( 11 ) is employed if the level of the good nearby pixels is relatively small (NS).

11. A method according to claim 9 , wherein average correction techniques ( 8 ) are employed if the level of the good nearby pixels is relatively large (NL).

12. Apparatus for correcting defects in image data comprising an array of pixels, the apparatus comprising:

means ( 5 ) for sampling the intensity of pixels of the image data in each side of the defect;

means ( 6 ) for calculating differences between the sampled intensities to generate one or more first intensity difference signals (D 1 , D 2 ) indicative of intensity differences across the defect;

means ( 6 ) for calculating differences between the sampled intensities to generate at least two second intensity difference signals (D 3 , D 4 ) indicative of an intensity difference on each respective side of the defect; and

means ( 7 - 11 ) for correcting the defect in dependence on the intensity difference signals (D 1 -D 4 ), a defect being replaced by an average of surrounding pixels if said first and second difference signals do not exceed a predetermined reference value (L 1 ,L 2 ).

13. A camera, comprising:

an image sensor (S 4 ) for furnishing image data; and

an apparatus for correcting defects in the image data as claimed in claim 12 .