Optical 3D digitizer with enlarged non-ambiguity zone

• US 20050018209A1
• Filed: 07/23/2004
• Published: 01/27/2005
• Est. Priority Date: 07/24/2003
• Status: Active Grant

• Current Assignee: Creaform, Inc. (Ametek Incorporated)
• Original Assignee: InSpeck, Inc. (Ametek Incorporated)
• Inventors: Rouleau, Patryck, Beauchamp, Dominique, Song, Li, Jacques, David, Glasson, Emmanuel, Lemelin, Guylain, Dufour, Patrick
• Granted Patent: US 7,274,470 B2
• Time in Patent Office: Days
• US Class Current: 356/604
• CPC Class Codes: G01B 11/25 by projecting a pattern, e.g. one or more lines, moiré fringes on the object G01B11/255 takes precedence ; image analysis for depth or shape recovery G06T7/50
• CPC Class Codes: G01B 11/2545 with one projection direction and several detection directions, e.g. stereo
• CPC Class Codes: G06T 7/521 from laser ranging, e.g. using interferometry; from the projection of structured light

Abstract

An optical 3D digitizer with an enlarged non-ambiguity zone, comprising a structured light projector for projecting a fringe pattern over a target area, the fringe pattern having a shiftable position over the target area is disclosed. First and second cameras having overlapping measurement fields are directed toward the target area and positioned with respect to the projector to define distinct triangulation planes therewith. The second camera has a larger non-ambiguity depth than the first camera. A computer evaluates a same set of camera-projector related functions from images captured by the cameras including the projected pattern at shifted positions, builds low depth resolution and degenerated 3D models from the camera-projector related functions evaluated with respect to the second and first cameras respectively, determines chromatic texture from the images, and builds a complete textured 3D model from data corresponding between the low depth resolution and degenerated 3D models within a tolerance range.

Claims

1. An optical 3D digitizer with an enlarged non-ambiguity zone, comprising:

at least one structured light projector for projecting a fringe pattern over a target area, the fringe pattern having a shiftable position over the target area;

a first camera directed toward the target area and positioned with respect to said at least one structured light projector to define a first triangulation plane therewith;

a second camera directed toward the target area and positioned with respect to said at least one structured light projector to define a second triangulation plane therewith, the second triangulation plane being distinct from the first triangulation plane, the first and second cameras having at least partially overlapping measurement fields, the second camera having a larger non-ambiguity depth than the first camera; and

a computer means connected to the cameras, for performing an image processing of images captured by the cameras, the image processing including evaluating a same set of camera-projector related functions from images including the pattern projected by said at least one structured light projector at shifted positions as captured by the cameras, building a low depth resolution 3D model from the camera-projector related functions evaluated with respect to the second camera, building a degenerated 3D model from the camera-projector related functions evaluated with respect to the first camera, determining chromatic texture from the images captured by the cameras, and building a complete textured 3D model from data corresponding between the low depth resolution and degenerated 3D models within a tolerance range.

2. The optical 3D digitizer according to claim 1, wherein the fringe pattern has a periodic sinusoidal section profile.

3. The optical 3D digitizer according to claim 1, wherein the set of camera-projector related functions comprises relations determined from phase shifting algorithms used in phase shifting interferometry.

4. The optical 3D digitizer according to claim 1, wherein the set of camera-projector related functions comprises relations determined from spatial phase shifting algorithms relying on a single fringe image in temporal phase shifting interferometry.

5. The optical 3D digitizer according to claim 1, wherein the images captured by the cameras including the pattern projected by said at least one structured light projector are written as:

I_n(i,j)=I_Ave(i,j)+I_Mod(i,j)[1+cos(φ

(i,j)+m(i,j)2π

+(n−

1)α

)]

where I represents one of the images, n represents a shift position of the pattern, i, j represent pixel coordinates in said one of the images, I_Ave(i, j) represents a local intensity diffused or reflected back to the camera which captured said one of the images, including surrounding ambient light and local temporal average intensity of the pattern, I_Mod(i, j) represents a local amplitude modulation of the pattern, φ

(i,j) represents a local phase function wrapped over 2π

range, m(i, j) represents a local order of the phase function, and α

represents a 90°

phase shift.

6. The optical 3D digitizer according to claim 1, wherein the camera-projector related functions comprise a phase function, a phase-shift function, an amplitude modulation function and an average function.

7. The optical 3D digitizer according to claim 6, wherein the computer means evaluates the chromatic texture from the amplitude modulation function and the average function using separate color channels.

8. The optical 3D digitizer according to claim 6, wherein the computer means evaluates the set of camera-projector related functions using:

$\begin{array}{c}\mathrm{\phi <br><br>}<br><br>\left(i,j\right)=\mathrm{inv}<br><br>\text{\hspace{1em}<br><br>}<br><br>\tan <br><br>\left(\left(I_4<br><br>\left(i,j\right)-I_2<br><br>\left(i,j\right)\right)/\left(I_1<br><br>\left(i,j\right)-I_3<br><br>\left(i,j\right)\right)\right)\\ I_{\mathrm{Mod}}<br><br>\left(i,j\right)=\sqrt{{\left(I_1<br><br>\left(i,j\right)-I_3<br><br>\left(i,j\right)\right)}^2+{\left(I_4<br><br>\left(i,j\right)-I_2<br><br>\left(i,j\right)\right)}^2}\\ I_{\mathrm{Ave}}<br><br>\left(i,j\right)=\left(I_1<br><br>\left(i,j\right)+I_2<br><br>\left(i,j\right)+I_3<br><br>\left(i,j\right)+I_4<br><br>\left(i,j\right)\right)/4\\ \mathrm{\alpha <br><br>}<br><br>\left(i,j\right)=2·<br><br>\mathrm{inv}<br><br>\text{\hspace{1em}<br><br>}<br><br>\tan <br><br>\left(\sqrt{\frac{3<br><br>\left(I_2<br><br>\left(i,j\right)-I_3<br><br>\left(i,j\right)-I_1<br><br>\left(i,j\right)-I_4<br><br>\left(i,j\right)\right)}{\left(I_2<br><br>\left(i,j\right)-I_3<br><br>\left(i,j\right)+I_1<br><br>\left(i,j\right)-I_4<br><br>\left(i,j\right)\right)}}\right)\end{array}$

where I represents one of the images, n represents a shift position of the pattern, i, j represent pixel coordinates in said one of the images, I_Ave (i, j) represents a local intensity diffused or reflected back to the camera which captured said one of the images, including surrounding ambient light and local temporal average intensity of the pattern, I_Mod(i, j) represents a local amplitude modulation of the pattern, φ

(i,j) represents a local phase function wrapped over 2π

range, and α

represents a phase shift; and

wherein a local order of the phase function m(i, j) is evaluated directly.

9. The optical 3D digitizer according to claim 8, wherein the computer means builds the low depth resolution and degenerated 3D models using look-up tables.

10. The optical 3D digitizer according to claim 1, wherein the image processing includes comparing the camera-projector related functions of the cameras and pixel coordinates in the images captured by the cameras, determining rejection of pixels without correspondences if a field of view of the second camera is entirely covered by the first camera, attaching pixels showing acceptable functions to surrounding pixels by local phase unwrapping, and extracting 3D coordinates forming the complete textured 3D model using table data for building the low depth resolution 3D model applied to corrected order values and current phase values.

11. An optical 3D digitizing method, comprising:

controllably projecting a fringe pattern over a target area using at least one structured light projector, the fringe pattern having a shiftable position over the target area;

positioning a first camera directed toward the target area with respect to said at least one structured light projector to define a first triangulation plane therewith;

positioning a second camera directed toward the target area with respect to said at least one structured light projector to define a second triangulation plane therewith, the second triangulation plane being distinct from the first triangulation plane, the first and second cameras having at least partially overlapping measurement fields, the second camera having a larger non-ambiguity depth than the first camera; and

performing an image processing of images captured by the cameras, the image processing including evaluating a same set of camera-projector related functions from images including the pattern projected by said at least one structured light projector at shifted positions as captured by the cameras, building a low depth resolution 3D model from the camera-projector related functions evaluated with respect to the second camera, building a degenerated 3D model from the camera-projector related functions evaluated with respect to the first camera, determining chromatic texture from the images captured by the cameras, and building a complete textured 3D model from data corresponding between the low depth resolution and degenerated 3D models within a tolerance range.

12. The optical 3D digitizing method according to claim 11, wherein the fringe pattern has a periodic sinusoidal section profile.

13. The optical 3D digitizing method according to claim 11, wherein the set of camera-projector related functions comprises relations determined from phase shifting algorithms used in phase shifting interferometry.

14. The optical 3D digitizing method according to claim 11, wherein the set of camera-projector related functions comprises relations determined from spatial phase shifting algorithms relying on a single fringe image in temporal phase shifting interferometry.

15. The optical 3D digitizing method according to claim 11, wherein the images captured by the cameras including the pattern projected by said at least one structured light projector are written as:

I_n(i,j)=I_Ave(i,j)+I_Mod(i,j)[1+cos(φ

(i,j)+m(i,j)2π

+(n−

1)]

where I represents one of the images, n represents a shift position of the pattern, i, j represent pixel coordinates in said one of the images, I_Ave(i, j) represents a local intensity diffused or reflected back to the camera which captured said one of the images, including surrounding ambient light and local temporal average intensity of the pattern, I_Mod(i, j) represents a local amplitude modulation of the pattern, φ

(i,j) represents a local phase function wrapped over 2π

range, m(i,j) represents a local order of the phase function, and α

represents a 90°

phase shift.

16. The optical 3D digitizing method according to claim 11, wherein the camera-projector related functions comprise a phase function, a phase-shift function, an amplitude modulation function and an average function.

17. The optical 3D digitizing method according to claim 16, wherein the chromatic texture is evaluated from the amplitude modulation function and the average function using separate color channels.

18. The optical 3D digitizing method according to claim 16, wherein the camera-projector related functions are evaluated using:

$\begin{array}{c}\mathrm{\phi <br><br>}<br><br>\left(i,j\right)=\mathrm{inv}<br><br>\text{\hspace{1em}<br><br>}<br><br>\tan <br><br>\left(\left(I_4<br><br>\left(i,j\right)-I_2<br><br>\left(i,j\right)\right)/\left(I_1<br><br>\left(i,j\right)-I_3<br><br>\left(i,j\right)\right)\right)\\ I_{\mathrm{Mod}}<br><br>\left(i,j\right)=\sqrt{{\left(I_1<br><br>\left(i,j\right)-I_3<br><br>\left(i,j\right)\right)}^2+{\left(I_4<br><br>\left(i,j\right)-I_2<br><br>\left(i,j\right)\right)}^2}\\ I_{\mathrm{Ave}}<br><br>\left(i,j\right)=\left(I_1<br><br>\left(i,j\right)+I_2<br><br>\left(i,j\right)+I_3<br><br>\left(i,j\right)+I_4<br><br>\left(i,j\right)\right)/4\\ \mathrm{\alpha <br><br>}<br><br>\left(i,j\right)=2·<br><br>\mathrm{inv}<br><br>\text{\hspace{1em}<br><br>}<br><br>\tan <br><br>\left(\sqrt{\frac{3<br><br>\left(I_2<br><br>\left(i,j\right)-I_3<br><br>\left(i,j\right)-I_1<br><br>\left(i,j\right)-I_4<br><br>\left(i,j\right)\right)}{\left(I_2<br><br>\left(i,j\right)-I_3<br><br>\left(i,j\right)+I_1<br><br>\left(i,j\right)-I_4<br><br>\left(i,j\right)\right)}}\right)\end{array}$

where I represents one of the images, n represents a shift position of the pattern, i, j represent pixel coordinates in said one of the images, I_Ave(i, j) represents a local intensity diffused or reflected back to the camera which captured said one of the images, including surrounding ambient light and local temporal average intensity of the pattern, I_Mod(i, j) represents a local amplitude modulation of the pattern, φ

(i,j) represents a local phase function wrapped over 2π

range, and α

represents a phase shift; and

wherein a local order of the phase function m(i, j) is evaluated directly.

19. The optical 3D digitizing method according to claim 18, wherein the low depth resolution and degenerated 3D models are built using look-up tables.

20. The optical 3D digitizing method according to claim 11, wherein the image processing includes comparing the camera-projector related functions of the cameras and pixel coordinates in the images captured by the cameras, determining rejection of pixels without correspondences if a field of view of the second camera is entirely covered by the first camera, attaching pixels showing acceptable functions to surrounding pixels by local phase unwrapping, and extracting 3D coordinates forming the complete textured 3D model using table data for building the low depth resolution 3D model applied to corrected order values and current phase values.

21. A computer apparatus for performing an image processing of images captured by first and second cameras, the second camera having a larger non-ambiguity depth than the first camera, comprising:

means for evaluating a same set of camera-projector related functions from images captured by the cameras, at least some of the images including a pattern projected at shifted positions;

means for building a low depth resolution 3D model from the camera-projector related functions evaluated with respect to the second camera;

means for building a degenerated 3D model from the camera-projector related functions evaluated with respect to the first camera;

means for determining chromatic texture from the images captured by the cameras; and

means for building a complete textured 3D model from data corresponding between the low depth resolution and degenerated 3D models within a tolerance range.

22. A computer readable medium having recorded thereon statements and instructions for execution by a computer to perform an image processing of images captured by first and second cameras directed toward a target area, the second camera having a larger non-ambiguity depth than the first camera, the image processing including evaluating a same set of camera-projector related functions from the images captured by the cameras, at least some of the images including a pattern projected at shifted positions, building a low depth resolution 3D model from the camera-projector related functions evaluated with respect to the second camera, building a degenerated 3D model from the camera-projector related functions evaluated with respect to the first camera, determining chromatic texture from the images captured by the cameras, and building a complete textured 3D model from data corresponding between the low depth resolution and degenerated 3D models within a tolerance range.

23. A computer program product, comprising a memory having computer readable code embodied therein, for execution by a CPU, for performing an image processing of images captured by first and second cameras directed toward a target area, the second camera having a larger non-ambiguity depth than the first camera, said code comprising:

code means for evaluating a same set of camera-projector related functions from the images captured by the cameras, at least some of the images including a pattern projected at shifted positions;

code means for building a low depth resolution 3D model from the camera-projector related functions evaluated with respect to the second camera;

code means for building a degenerated 3D model from the camera-projector related functions evaluated with respect to the first camera;

code means for determining chromatic texture from the images captured by the cameras; and

code means for building a complete textured 3D model from data corresponding between the low depth resolution and degenerated 3D models within a tolerance range.

24. A carrier wave embodying a computer data signal representing sequences of statements and instructions which, when executed by a processor, cause the processor to perform an image processing of images captured by first and second cameras directed toward a target area, the second camera having a larger non-ambiguity depth than the first camera, the statements and instructions comprising:

evaluating a same set of camera-projector related functions from the images captured by the cameras, at least some of the images including a pattern projected at shifted positions;

building a low depth resolution 3D model from the camera-projector related functions evaluated with respect to the second camera;

building a degenerated 3D model from the camera-projector related functions evaluated with respect to the first camera;

determining chromatic texture from the images captured by the cameras; and

building a complete textured 3D model from data corresponding between the low depth resolution and degenerated 3D models within a tolerance range.

25. An optical 3D digitizing method, comprising:

controllably projecting a fringe pattern having a shiftable position over a target area;

capturing images obtained by high depth resolution sensing and low depth resolution sensing from respective measurement fields at least partially overlapping each other over the target area;

determining absolute pixel 3D positions in the images obtained by low depth resolution sensing and high depth resolution sensing as a function of relations depending on the fringe pattern in the captured images and correspondence between the absolute pixel 3D positions in the images;

extracting chromatic texture from the captured images; and

building a complete textured 3D model from the absolute pixel 3D positions and the chromatic texture.

26. An optical 3D digitizer with an enlarged non-ambiguity zone, comprising:

at least one structured light projector projecting a fringe pattern over a target area, the fringe pattern having a shiftable position over the target area;

a first camera directed toward the target area and positioned with respect to said at least one structured light projector to define a first triangulation plane therewith;

a second camera directed toward the target area and positioned with respect to said at least one structured light projector to define a second triangulation plane therewith, the second triangulation plane being distinct from the first triangulation plane, the first and second cameras having at least partially overlapping measurement fields, the second camera having a larger non-ambiguity depth than the first camera; and

a computing device, connected to the cameras, being programmatically configured to i) evaluate a same set of camera-projector related functions from images including the pattern projected by said at least one structured light projector at shifted positions as captured by the cameras, ii) build a low depth resolution 3D model from the camera-projector related functions evaluated with respect to the second camera, iii) build a degenerated 3D model from the camera-projector related functions evaluated with respect to the first camera, iv) determine chromatic texture from the images captured by the cameras, and v) build a complete textured 3D model from data corresponding between the low depth resolution and degenerated 3D models within a tolerance range.

27. A computing device for performing image processing of images captured by first and second cameras, the second camera having a larger non-ambiguity depth than the first camera, wherein the computing device includes a software program which when executed in the computing device i) evaluates a same set of camera-projector related functions from images captured by the cameras, at least some of the images including a pattern projected at shifted positions, ii) builds a low depth resolution 3D model from the camera-projector related functions evaluated with respect to the second camera, iii) builds a degenerated 3D model from the camera-projector related functions evaluated with respect to the first camera, iv) determines chromatic texture from the images captured by the cameras, and v) builds a complete textured 3D model from data corresponding between the low depth resolution and degenerated 3D models within a tolerance range.