Method and apparatus for minimizing blurring and generating a high resolution image in a radiation imaging system

• US 6,166,384 A
• Filed: 11/06/1998
• Issued: 12/26/2000
• Est. Priority Date: 11/06/1998
• Status: Expired due to Term

• Current Assignee: General Electric Company
• Original Assignee: General Electric Company
• Inventors: Dentinger, Aaron Mark, Sowelam, Sameh Mahmoud, Han, Sung Su
• Primary Examiner(s): Hannaher, Constantine
• Assistant Examiner(s): GAGLIARDI, ALBERT J
• Application Number:
  US09/187,208
• Time in Patent Office: 781 Days
• Field of Search: 250/370.09, 250/370.11, 250/370.08, 250/371
• US Class Current: 250/370.09
• CPC Class Codes: G06T 5/20 using local operators
• CPC Class Codes: G06T 5/70 Denoising; Smoothing
• CPC Class Codes: G06T 5/73 Deblurring; Sharpening

Abstract

A method for generating a high resolution image in a radiation imaging system which produces multiple exposure signals from a movable detector array is disclosed. The image data signals or image acquisition signals generated from the image detector assembly of the image radiation system during an imaging cycle are passed through a finite impulse response filter having a combined frequency response comprised of an inverse filter frequency response portion and a noise regularization frequency response portion such that the noise regularization decrease the frequency response of the filter as the frequency of the data signals increase thereby effectively preventing blurring of the image produced from the image data signals.

Claims

1. A method of generating a high resolution image of incident radiation by a radiation imaging system which produces multiple exposure image data signals from a movable detector array comprising passing the image data signals through a finite impulse response deblurring filter having a combined frequency response comprising an inverse filter frequency response portion and a noise regularization frequency response portion wherein the noise regularization decreases the frequency response of the filter as the frequency of the signal increases.

2. The method according to claim 1 wherein the combined frequency response of the filter is given by:

##EQU5## wherein;

H(v_r)=combined frequency response H_inv (v_r)=inverse filter frequency response

H_reg (v_r)=regulation frequency response ##EQU6## v_x, v_y =spatial frequencies in x and y directions α

=transition frequency between inversion and regularization portions.

3. The method according to claim 2 further comprising selecting a regularization function of the filter as an n^th order polynomial wherein:

H_reg (α

)=H_inv (α

)

H_inv =inverse filter frequency response

H_reg =regularization frequency response

α

=transition frequency between inverse filter frequency response portion and noise regularization frequency response portion.

4. The method according to claim 2 further comprising selecting a regularization function of the filter as an n^th order polynomial wherein:

##EQU7## H_reg =regularization frequency response H_inv =inverse filter frequency response ##EQU8## v_x, v_y =spatial frequencies in x and y directions α

=transition frequency between inverse filter frequency response portion and noise regularization frequency response portion.

5. The method according to claim 1 wherein image data signals produced by the radiation imaging system have an oversampled Nyquist frequency and further comprising selecting the noise regularization frequency response portion of the filter as an n^th order polynomial having zero response at the oversampled Nyquist frequency wherein:

H_reg (v_NYQUIST)=0

where;

H_reg (v_NYQUIST)=noise regularization frequency response at Nyquist frequency.

6. The method according to claim 1 wherein the image data signals produced by the radiation imaging system have an oversampled Nyquist frequency and further comprising selecting the noise regularization frequency response portion of the filter as an n^th order polynomial having zero slope at the Nyquist frequency wherein:

##EQU9## where;

H_reg (v_NYQUIST)=noise regularization frequency response at Nyquist frequency ##EQU10## v_x, v_y =spatial frequency in x and y directions.

7. The method according to claim 1 wherein the image data signals produced by the radiation imaging system have an oversampled Nyquist frequency and, a frequency α

at the transition between the inverse frequency response portion and the noise regularization frequency response portion, and further comprising selecting the noise regularization frequency response portion of the filter as an n^th order polynomial having;

a) continuity at α

;

b) a continuous slope at α

;

c) zero response at the oversampled Nyquist frequency; and

d) zero slope at the oversampled Nyquist frequency.

8. The method according to claim 2 wherein image data signals produced by the radiation imaging system have an oversampled Nyquist frequency and further comprising selecting the noise regularization frequency response portion of the filter as an n^th order polynomial having zero response at the oversampled Nyquist frequency wherein:

H_reg (v_NYQUIST)=0

where;

H_reg (v_NYQUIST)=noise regularization frequency response at Nyquist frequency.

9. The method according to claim 2 wherein the image data signals produced by the radiation imaging system have an oversampled Nyquist frequency and further comprising selecting the noise regularization frequency response portion of the filter as an n^th order polynomial having zero slope at the Nyquist frequency wherein:

##EQU11## where;

H_reg (v_NYQUIST)=noise regularization frequency response at Nyquist frequency ##EQU12## v_x, v_y =spatial frequency in x and y directions.

10. The method according to claim 2 wherein the image data s signals produced by the radiation imaging system have an oversampled Nyquist frequency and, a frequency α

at the transition between the inverse frequency response portion and the noise regularization n frequency response portion, and further comprising selecting the noise regularization frequency response portion of the filter as an n^th order polynomial having;

a) continuity at α

;

b) a continuous slope at α

;

c) zero response at the oversampled Nyquist frequency; and

d) zero slope at the oversampled Nyquist frequency.

11. A method of generating a high resolution imaging of incident radiation in a radiation imaging system, comprising the steps of:

sequentially positioning an array of photosensor pixels in first through K^th imaging positions in accordance with a predetermined imaging cycle, said photosensor pixels being disposed in said array in rows and columns having a predetermined photosensor pixel pitch therebetween which determines the imaging system'"'"'s Nyquist frequency, each of said sequential imaging positions being selected such that the area imaged by each respective pixel overlaps a portion of the area imaged by said respective pixel while disposed in the preceding imaging position in said predetermined imaging cycle so as to oversample the imaged area;

storing the respective image data signals generated in each of the first through K^th imaging positions; and

deblurring the image data signals by passing said signals through a finite impulse response filter having a combined frequency response portion and a noise regularization frequency response portion wherein the noise regularization portion serves to decrease the frequency response of the filter as the frequency of the image data signals increase to thereby develop said high resolution image.

12. The method according to claim 11 wherein the combined frequency response of the filter is given by:

##EQU13## wherein;

H(v_r)=combined frequency response H_inv (v_r)=inverse filter frequency response

H_reg (v_r)=regulation frequency response ##EQU14## v_x, v_y =spatial frequencies in x and y directions α

=transition frequency between inversion and regularization portions.

13. The method according to claim 12 further comprising selecting a regularization function of the filter as an n^th order polynomial wherein:

H_reg (α

)=H_inv (α

)

H_inv =inverse filter frequency response

H_reg =regularization frequency response

α

=transition frequency between inverse filter frequency response portion and noise regularization frequency response portion.

14. The method according to claim 12 further comprising selecting a regularization function of the filter as an n^th order polynomial wherein:

##EQU15## H_reg =regularization frequency response H_inv =inverse filter frequency response ##EQU16## v_x, v_y =spatial frequencies in x and y directions α

=transition frequency between inverse filter frequency response portion and noise regularization frequency response portion.

15. The method according to claim 11 wherein image data signals produced by the radiation imaging system have an oversampled Nyquist frequency and further comprising selecting the noise regularization frequency response portion of the filter as an n^th order polynomial having zero response at the oversampled Nyquist frequency wherein:

H_reg (v_NYQUIST)=0

where;

H_reg (v_NYQUIST)=noise regularization frequency response at Nyquist frequency.

16. The method according to claim 11 wherein the image data signals produced by the radiation imaging system have an oversampled Nyquist frequency and further comprising selecting the noise regularization frequency response portion of the filter as an n^th order polynomial having zero slope at the Nyquist frequency wherein:

##EQU17## where;

H_reg (v_NYQUIST)=noise regularization frequency response at Nyquist frequency ##EQU18## vx, v_y =spatial frequency in x and y directions.

17. The method according to claim 11 wherein the image data signals produced by the radiation imaging system have an oversampled Nyquist frequency and, a frequency α

at the transition between the inverse frequency response portion and the noise regularization frequency response portion, and further comprising selecting the noise regularization frequency response portion of the filter as an n^th order polynomial having;

a) continuity at α

;

b) a continuous slope at α

;

c) zero response at the oversampled Nyquist frequency; and

d) zero slope at the oversampled Nyquist frequency.

18. The method according to claim 12 wherein image data signals produced by the radiation imaging system have an oversampled Nyquist frequency and further comprising selecting the noise regularization frequency response portion of the filter as an n^th order polynomial having zero response at the oversampled Nyquist frequency wherein:

H_reg (v_NYQUIST)=0

where;

H_reg (v_NYQUIST)=noise regularization frequency response at Nyquist frequency.

19. The method according to claim 12 wherein the image data signals produced by the radiation imaging system have an oversampled Nyquist frequency and further comprising selecting the noise regularization frequency response portion of the filter as an n^th order polynomial having zero slope at the Nyquist frequency wherein:

##EQU19## where;

H_reg (v_NYQUIST)=noise regularization frequency response at Nyquist frequency ##EQU20## v_x, v_y =spatial frequency in x and y directions.

20. The method according to claim 12 wherein the image data signals produced by the radiation imaging system have an oversampled Nyquist frequency and, a frequency α

at the transition between the inverse frequency response portion and the noise regularization frequency response portion, and further comprising selecting the noise regularization frequency response portion of the filter as an n^th order polynomial having;

a) continuity at α

;

b) a continuous slope at α

;

c) zero response at the oversampled Nyquist frequency; and

d) zero slope at the oversampled Nyquist frequency.

21. The method of claim 11 wherein the step of sequentially positioning said array of photosensor pixels further comprises the step of displacing said photosensor pixel array in a selected sequence so that said array is respectively disposed at each of four imaging positions during said imaging cycle.

22. The method of claim 21 wherein the step of displacing said array further comprises sequentially translating said array along a first axis and a second axis, said first and second axes being orthogonal.

23. The method of claim 22 wherein the distance along said first and second axes between respective ones of said four imaging positions correspond to one-half of photosensor pixel pitch.

24. The method of claim 23 further comprising the step of driving a display and analysis module to present said high resolution data.

25. A radiation imaging system comprising:

a photosensor pixel array, said array adapted to generate an image data signal being sequentially disposed in selected ones of a plurality of sequential imaging positions pursuant to a predetermined imaging cycle, each of said sequential imaging positions being selected so as to oversample the imaged area; and

an image processor coupled to said photosensor pixel array so as to receive the respective image data signals generated, said image processor being adapted to store said respective image data signals generated during an imaging cycle in each of said respective sequential imaging positions as an unfiltered data set, said image processor further comprising a finite impulse response deblurring filter adapted to selectively filter said data set to generate a non-aliased fine resolution data set having spatial resolution greater than the imaging system Nyquist frequency determined by pixel pitch of said photosensor, array, said deblurring filter having a combined frequency response comprising an inverse filter frequency response portion and a noise regularization frequency response portion wherein the noise regularization decreases the frequency response of the filter as the frequency of the signal increases.

26. An imaging system according to claim 25 wherein the combined frequency response of the filter is given by:

##EQU21## wherein;

H(v_r)=combined frequency response

H_inv (v_r)=inverse filter frequency response

H_reg (v_r)=regulation frequency response ##EQU22## v_x, v_y =spatial frequencies in x and y directions α

=transition frequency between inversion and regulariztion portions.

27. An imaging system according to claim 26 wherein the regularization function of the filter in an n^th order polynomial wherein:

H_reg (α

)=H_inv (α

)

H_inv =inverse filter frequency response

H_reg =regularization frequency response

α

=transition frequency between inverse filter frequency response portion and noise regularization frequency response portion.

28. In imagine system according to claim 26 wherein the regularization function of the filter is an nt order polynomial wherein:

##EQU23## H_reg =regularization frequency response H_inv =inverse filter frequency response ##EQU24## v_x, v_y =spatial frequencies in x and y directions α

=transition frequency between inverse filter frequency response portion and noise regularization frequency response portion.

29. An imaging system according to claim 25 wherein the noise regularization frequency response portion of the filter is an n^th order polynomial having zero response at the oversampled Nyquist frequency wherein:

H_reg (v_NYOUIST)=0

where;

H_reg (v_NYQUIST)=noise regularization frequency response at Nyquist frequency.

30. An imaging system according to claim 25 wherein he image data signals produced by the radiation imaging system have an oversampled Nyquist frequency and wherein the noise regularization frequency response portion of the filter is an n^th order polynomial having zero slope at the Nyquist frequency wherein:

##EQU25## where;

H_reg (v_NYQUIST)=noise regularization frequency response at Nquist frequency ##EQU26## v_x, v_y =spatial frequency in x and y directions.

31. An imaging system according to claim 25 wherein the image data signals produced by the radiation imaging system have an oversampled Nyquist frequency and, a frequency α

at the transition between the inverse frequency response portion and the noise regularization frequency response portion of the combined frequency response of the filter and the noise regularization frequency response portion of the filter is an n^th order polynomial having;

a) continuity at α

;

b) a continuous slope at α

;

c) zero response at the oversampled Nyquist frequency; and

d) zero slope at the oversampled Nyquist frequency.

32. An imaging system according to claim 26 wherein the image data signals produced by the radiation imaging system have an oversampled Nyquist frequency and wherein the noise regularization frequency response portion of the filter as an n^th order polynomial having zero response at the oversampled Nyquist frequency wherein:

H_reg (v_NYQUIST)=0

where;

H_reg (v_NYQUIST)=noise regularization frequency response at Nyquist frequency.

33. An imaging system according to claim 26 wherein the image data signals produced by the radiation imaging system have an oversampled Nyquist frequency and wherein the noise regularization frequency response portion of the filter is an n^th order polynomial having zero slope at the Nyquist frequency wherein:

##EQU27## where;

H_reg (v_NYQUIST)=noise regularization frequency response at Nyquist frequency ##EQU28## v_x, v_y =spatial frequency in x and y directions.

34. An imaging system according to claim 26 wherein the image data signals produced by the radiation imaging system have an oversampled Nyquist frequency and, a frequency α

at the transition between the inverse frequency response portion and the noise regularization frequency response portion, and wherein the noise regularization frequency response portion of the filter is an n^th order polynomial having;

a) continuity at α

;

b) a continuous slope at α

;

c) zero response at the oversampled Nyquist frequency; and

d) zero slope at the oversampled Nyquist frequency.