System for generating thermographic images using thermographic signal reconstruction

• US 7,724,925 B2
• Filed: 05/18/2004
• Issued: 05/25/2010
• Est. Priority Date: 12/02/1999
• Status: Active Grant

• Current Assignee: Thermal Wave Imaging, Inc.
• Original Assignee: Thermal Wave Imaging, Inc.
• Inventors: Shepard, Steven M.
• Primary Examiner(s): Le; Brian Q
• Application Number:
  US10/848,274
• Publication Number:
  US 20050008215A1
• Time in Patent Office: 2,198 Days
• Field of Search: 382/115, 382/116, 382/124, 382/125, 382/141, 382/143, 382/149, 382/152, 348/86, 348/125, 250/328, 250/329, 250/330, 250/332, 250/333, 250/334, 250/358.1, 250/341.6, 250/339.02, 250/339.14, 250/341.1, 374/5, 374/10, 374/120, 374/121, 374/124, 700/95, 700/212, 356/51, 209/577, 600/300
• US Class Current: 382/115
• CPC Class Codes: G01N 25/72 Investigating presence of flaws

Abstract

A method for compiling thermographic data including obtaining data indicative of a monotonically changing characteristic of a specimen, sequencing the data or a surrogate of the data into a plurality of groups, categorizing, within each group, the frequency distribution of an attribute of the data or an attribute of said surrogate data, and compiling, from one or more groups, a collection of two or more of the frequency distributions.

Claims

1. A method of compiling thermographic data, comprising the steps of:

A) using a thermographic camera to obtain a thermographic image data indicative of a monotonically changing, thermal characteristic of a specimen,

B) using an image processor to sequence the thermographic image data or a surrogate of the thermographic image data into a plurality of time based, thermographic image data frames,

C) categorizing, for each frame, a frequency distribution of frame pixels, wherein each pixel is encoded with a thermographic image data value or a surrogate of a thermographic image data value,

D) compiling, from two or more frames, a collection of the frequency distributions categorized in step C).

2. The method of claim 1, further including the step of:

E) comparing the compiled frequency distributions of step D) to a predetermined set of compiled frequency distributions.

3. The method of claim 2, wherein the comparing step includes the sub-step of calculating a correlation coefficient based on the correlation between the compiled frequency distributions of step D) to the compiled frequency distributions of step E).

4. The method of claim 3, wherein the comparing step includes calculating a correction coefficient according to the following formula:

$r=\frac{\underset{M}{\overset{}{\sum <br><br>}}<br><br><br><br>\underset{N}{\overset{}{\sum <br><br>}}<br><br>\left(A_{\mathrm{MN}}-\stackrel{\_}{A}\right)<br><br>\left(B_{\mathrm{MN}}-\stackrel{\_}{B}\right)}{\sqrt{\left(\underset{M}{\overset{}{\sum <br><br>}}<br><br><br><br>\underset{N}{\overset{}{\sum <br><br>}}<br><br>{\left(A_{\mathrm{MN}}-\stackrel{\_}{A}\right)}^2\right)<br><br>\left(\underset{M}{\overset{}{\sum <br><br>}}<br><br><br><br>\underset{N}{\overset{}{\sum <br><br>}}<br><br>{\left(B_{\mathrm{MN}}-\stackrel{\_}{B}\right)}^2\right)}}$

wherein;

r=the measure of correlation between the “

gold standard”

fingerprint and the specimen fingerprint,

A_MN=the value of the row, column pixel from the “

gold standard”

fingerprint,

B_MN=the value of the row, column pixel of the specimen fingerprint,

Ā

represents the average value of all of the pixel values from the “

gold standard”

fingerprint,

B represents the average value of all of the pixel values from the sample fingerprint.

5. The method of claim 2, wherein said comparing step includes the step of calculating the difference between the frequency distributions of step D) and the frequency distributions of step E).

6. The method of claim 2, wherein said comparing step includes the step of dividing the frequency distributions of step D) by the frequency distribution of step E).

7. The method of claim 2, wherein said comparing step includes the step of dividing the frequency distributions of step E) by the frequency distributions of step D).

8. The method of claim 1, wherein obtaining data includes using a camera to obtain a plurality of time based infrared images.

9. The method of claim 1, wherein the thermographic image surrogate includes a polynomial fitted to the data.

10. The method of claim 9, wherein the data surrogate includes a first, second, or third derivative of said fitted polynomial.

11. The method of claim 1, wherein the attribute of the thermographic image of the data or said surrogate of the thermographic image data includes thermal energy or thermal intensity.

12. The method of claim 1, wherein the obtained thermographic image data is derived from raw data.

13. The method of claim 1, wherein the thermographic image data or the surrogate of the thermographic image data is derived from a first derivative of a log-log curve.

14. The method of claim 1, wherein the thermographic image data or the surrogate of the thermographic image data is derived from a second derivative of a log-log curve.

15. The method of claim 1, wherein the thermographic image data or the surrogate of the thermographic image data is derived from a curvature of a log-log curve.

16. A method of compiling thermographic data, comprising the steps of:

A) using an infrared camera to capture, at two or more distinct times (t_a, t_b, . . . t₂), two or more respectively associated frames (frame_a, frame_b, . . . frame_z) of thermographic data of a sample specimen, wherein the thermographic data is indicative of a monotonically changing, thermal characteristic of the sample specimen;

B) using an image processor to sequence the frames of thermographic data into a series of histograms (hist_a, hist_b, . . . hist_z) wherein each histogram is respectively associated with a frame (frame_a, frame_b, . . . frame_z) of thermographic data; and

C) compiling two or more of the of histograms to form a thermographic fingerprint of the sample specimen.

17. The method according to claim 16, further comprising the step of:

D) using an infrared camera to capture a gold standard thermographic fingerprint of a benchmark specimen, wherein the gold standard thermographic fingerprint of the benchmark specimen defines the benchmark specimen to have a defect-free attribute; and

E) comparing the thermographic fingerprint of the sample specimen to the gold standard fingerprint of the benchmark specimen for

F) determining the acceptability of the sample specimen.

18. The method according to claim 17, wherein prior to the compiling step, further comprising the step of:

B1) categorizing the thermographic data into

a first frequency distribution containing a first attribute of the thermographic data, and

a second frequency distribution containing a second attribute of the thermographic data.

19. The method according to claim 18, wherein the first attribute of the thermographic data is indicative of the sample specimen including:

a defect, wherein the second attribute of the thermographic data is indicative of the sample specimen being defect-free.

20. The method of claim 17, wherein the comparing step includes the sub-step of calculating a correlation coefficient.

21. The method of claim 20, wherein the calculating step includes the following formula:

$r=\frac{\underset{M}{\sum <br><br>}<br><br>\underset{N}{\sum <br><br>}<br><br>\left(A_{\mathrm{MN}}-\stackrel{\_}{A}\right)<br><br>\left(B_{\mathrm{MN}}-\stackrel{\_}{B}\right)}{\sqrt{\left(\underset{M}{\sum <br><br>}<br><br>\underset{N}{\sum <br><br>}<br><br>{\left(A_{\mathrm{MN}}-\stackrel{\_}{A}\right)}^2\right)<br><br>\left(\underset{M}{\sum <br><br>}<br><br>\underset{N}{\sum <br><br>}<br><br>{\left(B_{\mathrm{MN}}-\stackrel{\_}{B}\right)}^2\right)<br><br>}}$

wherein;

r=the measure of correlation between the gold standard thermographic fingerprint of the benchmark specimen and the fingerprint of the sample specimen,

A_MN=the value of the row, column pixel from the gold standard thermographic fingerprint,

B_MN=the value of the row, column pixel of the fingerprint of the sample specimen,

Ā

represents the average value of all of the pixel values from the gold standard thermographic fingerprint,

B represents the average value of all of the pixel values from the fingerprint of the sample specimen.

22. A method of compiling thermographic data, comprising the steps of:

A) exposing a specimen to an external thermal excitation event;

B) using an infrared camera to capture, over a first period of time, thermographic image data of a the specimen, wherein the thermographic image data is indicative of a monotonically changing, time based thermal characteristic of the specimen in response to the thermal event;

C) using an image processor to sequence the thermographic image data of the specimen into a time series of frequency distributions wherein each frequency distribution in said series of frequency distributions is respectively associated with a portion of the thermographic image data; and

D) compiling the series of frequency distributions to form a thermographic fingerprint of the specimen.

23. The method according to claim 22, further comprising the steps of:

E) wherein said step B) further includes obtaining two or more frames of thermographic image data of said specimen, wherein each frame of the thermographic image data of the specimen is indicative of a monotonically changing, time based thermal characteristic of the specimen;

F) wherein step C) further includes sequencing the thermographic image data frames of the specimen into a chronological time series;

G) compiling the chronological time series of data frames to form a thermographic fingerprint of the sample specimen; and

H) comparing the thermographic fingerprint of the specimen against a gold standard fingerprint of a benchmark specimen for

I) determining the acceptability of the specimen.

24. The method of claim 23, wherein the comparing step includes the sub-step of calculating a correlation coefficient.

25. The method of claim 24, wherein the calculating step includes the following formula:

$r=\frac{\underset{M}{\sum <br><br>}<br><br><br><br>\underset{N}{\sum <br><br>}<br><br><br><br>\left(A_{\mathrm{MN}}-\stackrel{\_}{A}\right)<br><br>\left(B_{\mathrm{MN}}-\stackrel{\_}{B}\right)}{\sqrt{\left(\underset{M}{\sum <br><br>}<br><br><br><br>\underset{N}{\sum <br><br>}<br><br><br><br>{\left(A_{\mathrm{MN}}-\stackrel{\_}{A}\right)}^2\right)<br><br>\left(\underset{M}{\sum <br><br>}<br><br><br><br>\underset{N}{\sum <br><br>}<br><br>{\left(B_{\mathrm{MN}}-\stackrel{\_}{B}\right)}^2\right)}}$

wherein;

r=the measure of correlation between the gold standard thermographic fingerprint of the benchmark specimen and the fingerprint of the sample specimen,

A_MN=the value of the row, column pixel from the gold standard thermographic fingerprint,

B_MN=the value of the row, column pixel of the fingerprint of the sample specimen,

Ā

represents the average value of all of the pixel values from the gold standard thermographic fingerprint,

B represents the average value of all of the pixel values from the fingerprint of the sample specimen.