Method and system for fast calibration of three-dimensional (3D) sensors

• US 7,471,376 B2
• Filed: 07/06/2007
• Issued: 12/30/2008
• Est. Priority Date: 07/06/2006
• Status: Active Grant

• Current Assignee: Microsoft Technology Licensing LLC (Microsoft Corporation)
• Original Assignee: Canesta Incorporated (Microsoft Corporation)
• Inventors: Yalcin, Hakan, Bamji, Cyrus
• Primary Examiner(s): Tarcza; Thomas H
• Assistant Examiner(s): Ratcliffe; Luke D
• Application Number:
  US11/825,582
• Publication Number:
  US 20080007709A1
• Time in Patent Office: 543 Days
• Field of Search: 356 301- 315, 356 401- 41, 356 501- 515, 356 6- 22
• US Class Current: 356/5.01
• CPC Class Codes: G01C 25/00 Manufacturing, calibrating, cleaning, or repairing instruments or devices referred to in the other groups of this subclass testing, calibrating or compensating compasses G01C17/38
• CPC Class Codes: G01C 3/08 Use of electric radiation detectors
• CPC Class Codes: G01S 17/36 with phase comparison between the received signal and the contemporaneously transmitted signal
• CPC Class Codes: G01S 17/894 3D imaging with simultaneous measurement of time-of-flight at a 2D array of receiver pixels, e.g. time-of-flight cameras or flash lidar
• CPC Class Codes: G01S 7/497 Means for monitoring or calibrating

Abstract

Rapid calibration of a TOF system uses a stationary target object and electrically introduces phase shift into the TOF system to emulate target object relocation. Relatively few parameters suffice to model a parameterized mathematical representation of the transfer function between measured phase and Z distance. The phase-vs-distance model is directly evaluated during actual run-time operation of the TOF system. Preferably modeling includes two components: electrical modeling of phase-vs-distance characteristics that depend upon electrical rather than geometric characteristics of the sensing system, and elliptical modeling that phase-vs-distance characteristics that depending upon geometric rather than electrical characteristics of the sensing system.

Claims

1. A method of calibrating a time-of-flight (TOF) system of the type that emits light of a known phase, detects a portion of said light reflected from a target object a distance Z away, and determines Z by examining phase shift in detected reflected light relative to said known phase of emitted light, the method comprising the following steps:

(a) disposing a target object a distance Z^x from said TOF system, said distance Z^x being within operating distance range of said TOF system;

(b) altering said known phase of said emitted light by at least two known phase values;

(c) for each known phase value of said emitted light, determining from detected reflected light a corresponding phase shift relative to said known phase;

(d) using corresponding relative phase shift determined at step (c) to form an electrical model of detection characteristics of said TOF system; and

(e) storing data representing said electrical model;

wherein data stored at step (e) is acquirable using only optical energy emitted by said TOF system and traversing said distance Z^x and is useable during run-time operation of said TOF system to provide calibrated values of Z responsive to phase shift in detected reflected light.

2. The method of claim 1, where said Z^x is a shortest distance Z^f whereat elliptical error arising from geometry of phase detection of said TOF system is negligible.

3. The method of claim 1, wherein step (b) includes sweeping said first phase with incremental values of phase having at least one characteristic selected from a group consisting of (i) increments between each of said phase values are equal in magnitude, (ii) increments between at least some of said phase values have different magnitude, and (iii) sweeping encompasses substantially a range of about 0°

to about 360°

.

4. The method of claim 1, wherein step (e) includes storing said data representing said electrical model within said TOF system.

5. The method of claim 1, wherein step (d) includes forming said electrical model as a parametric function characterized by at least two parameters.

6. The method of claim 1, wherein step (b) includes sweeping said known phase with incremental values of phase exceeding 360°

, and wherein step (d) includes unwrapping relative phase shift determined at step (c) to avoid distance ambiguity.

7. The method of claim 1, wherein said model formed at step (d) includes a linear factor and a sinusoid factor.

8. The method of claim 1, wherein:

said TOF system includes an array of detectors;

said model formed at step (d) approximates Z=ZUD_ij·

[p+m_ij+A_ij sin(s_ijp+2π

fp)], where at least two parameters of ZUD_ij, m_ij, A_ij, and s_ij are per detector parameters, f is a global TOF system parameter, and p is phase.

9. The method of claim 1, further including a step of dealiasing phase shift in said detected reflected light, whereby said electrical model formed at step (d) is useable during run-time operation of said TOF system for unwrapped phase exceeding 360°

.

10. The method of claim 1, wherein step (d) further includes forming an elliptical error model to correct phase-vs-distance data for geometric characteristics of said TOF system.

11. The method of claim 10, wherein said elliptical model formed at step (d) is useable when Z<

Z^f, where Z^f is a shortest distance at which elliptical error for said TOF system is negligible.

12. The method of claim 10, wherein forming an elliptical error model includes the following steps:

(i) disposing a target object at at least one distance Z^y<

Z^f from said TOF system, where Z^f is a shortest distance whereat elliptical error arising from geometry of phase detection of said TOF system is negligible;

(ii) for each said distance Z^y, determining from detected reflect light a corresponding phase shift relative to said known phase;

(iii) for each said distance Z^y, obtaining a phase value from said electrical model formed at step (d); and

(iv) obtaining a difference in phase value between phase determined at step (ii) and phase obtained from step (iii), and using said difference in phase to form an elliptical error model;

wherein said elliptical error model is useable during run-time operation of said TOF system to provide improved calibrated values of Z<

Z^f responsive to phase shift in detected reflected light.

13. The method of claim 12, wherein step (iv) includes forming said elliptical error model as a parametric function.

14. A method of improving elliptical error calibration in a time-of-flight (TOF) system of the type that emits light of a known phase, detects a portion of said light reflected from a target object a distance Z away, and determines Z by examining phase shift in detected reflected light relative to said known phase of emitted light, the method comprising the following steps:

(i) disposing a target object at at least one distance Z^y<

Z^f from said TOF system, where Z^f is a shortest distance whereat elliptical error arising from geometry of phase detection of said TOF system is negligible;

(ii) for each said distance Z^y, determining from detected reflect light a corresponding phase shift relative to said known phase;

(iii) for each said distance Z^y,obtaining a phase value from an electrical model of phase-vs-distance formed for said TOF system; and

(iv) obtaining a difference in phase value between phase determined at step (ii) and phase obtained from step (iii), and using said difference in phase to form an elliptical error model;

wherein said elliptical error model is acquirable using only optical energy emitted by said TOF system and traversing said at least one distance Z^y<

Z^f and is useable during run-time operation of said TOF system to provide improved calibrated values of Z <

Z^f responsive to phase shift in detected reflected light.

15. The method of claim 14, wherein step (iv) includes forming said elliptical error model as a parametric function.

16. The method of claim 14, wherein step (iv) includes storing said elliptical error model in memory useable by said TOF system during run-time operation of said TOF system.

17. A time-of-flight (TOF) system of the type that emits light of a known phase, detects a portion of said light reflected from a target object a distance Z away, and determines Z by examining phase shift in detected reflected light relative to said known phase of emitted light, the TOF including means for altering known phase emitted by said TOF system, and further including memory storing a distance-vs-phase calibration model used to calibrate said TOF system, said calibration model obtained according to a method comprising the following steps:

(a) disposing a target object a distance Z^x from said TOF system, said distance Z^x being within operating distance range of said TOF system;

(b) causing said means for altering known phase to vary said known phase of said emitted light by at least two known phase values;

(c) for each known phase value of said emitted light, determining from detected reflected light a corresponding phase shift relative to said known phase;

(d) using corresponding relative phase shift determined at step (c) to form an electrical model of detection characteristics of said TOF system; and

(e) storing data representing said electrical model in said memory;

wherein data stored in memory at step (e) is acquirable using only optical energy emitted by said TOF system and traversing said at least one distance Z^x and is useable during run-time operation of said TOF system to provide calibrated values of Z responsive to phase shift in detected reflected light.

18. The TOF system of claim 17, wherein at step (a), said Z^x is a shortest distance Z^f whereat elliptical error arising from geometry of phase detection of said TOF system is negligible.

19. The TOF system of claim 17, wherein said memory further includes an elliptical model of detection characteristics of said TOF system that are substantially independent of electrical characteristics, said elliptical model being used at distances Z<

Z^f, where Z^f is a shortest distance at which elliptical error is substantially negligible.

20. The TOF system of claim 19, wherein said elliptical model of detection characteristics stored in said memory is formed as follows:

(i) disposing a target object at at least one distance Z<sup>y<

Zf</sup> from said TOF system, where Z^f is a shortest distance whereat elliptical error arising from geometry of phase detection of said TOF system is negligible;

(ii) for each said distance Z^y, determine from detected reflect light a corresponding phase shift relative to said known phase;

(iii) for each said distance Z^y, obtain a phase value from said electrical model formed at step (d);

(iv) obtain a difference in phase value between phase determined at step (ii) and phase obtained from step (iii), and use said difference in phase to form an elliptical error model;

wherein said elliptical error model is useable during run-time operation of said TOF system to provide improved calibrated values of Z<

Z^f responsive to phase shift in detected reflected light.