Wavelet compression of gravity correction data

• US 6,651,002 B1
• Filed: 12/20/2002
• Issued: 11/18/2003
• Est. Priority Date: 12/20/2002
• Status: Expired due to Term

• Current Assignee: Honeywell International Inc.
• Original Assignee: Honeywell International Inc.
• Inventors: Morgan, Kenneth A.
• Primary Examiner(s): Cuchlinski, Jr., William A.
• Assistant Examiner(s): MANCHO, RONNIE M
• Application Number:
  US10/325,073
• Time in Patent Office: 333 Days
• Field of Search: 701/220, 701/217, 701/216, 701/214, 701/221, 701/224, 702/141, 342/357.13, 342/357.14
• US Class Current: 701/220
• CPC Class Codes: G01C 21/185 for gravity

Abstract

Compressed wavelet transform coefficients are used to correct outputs of inertial sensors for gravitational deflections. The compressed wavelet transform coefficients comprise gravitational deflection data that is first converted to discrete wavelet coefficients by a discrete wavelet transform and that are then compressed to form the compressed wavelet transform coefficients. An inertial sensor signal relating to a first navigational parameter of a vehicle is generated. The compressed discrete wavelet coefficients are accessed from a memory based on a position of the vehicle. The accessed coefficients are decompressed, and an inverse discrete wavelet transform is performed on the decompressed coefficients to produce gravitational compensation data. The sensor signal is compensated based on the gravitational compensation data.

Claims

1. A vehicle navigation method comprising:

generating an inertial sensor signal relating to a navigational parameter of a vehicle;

accessing discrete wavelet coefficients from a memory based on a position of the vehicle;

performing an inverse discrete wavelet transform on the discrete wavelet coefficients to produce compensation data; and

,

compensating the inertial sensor signal based on the compensation data.

2. The vehicle navigation method of claim 1 wherein the performing of an inverse discrete wavelet transform comprises:

performing the inverse discrete wavelet transform by column on the discrete wavelet coefficients to produce a matrix; and

,

performing the inverse discrete wavelet transform by row on the matrix to produce the compensation data.

3. The vehicle navigation method of claim 1 wherein the performing of an inverse discrete wavelet transform comprises:

scaling the discrete wavelet coefficients; and

,

performing an inverse discrete wavelet transform on the scaled coefficients to produce the compensation data.

4. The vehicle navigation method of claim 3 wherein the performing of an inverse discrete wavelet transform comprises:

performing the inverse discrete wavelet transform by column on the scaled coefficients to produce a matrix; and

,

performing the inverse discrete wavelet transform by row on the matrix to produce the compensation data.

5. The vehicle navigation method of claim 1 wherein the performing of an inverse discrete wavelet transform comprises:

up sampling and inverse low pass filtering the discrete wavelet coefficients to produce first up sampled and filtered data;

up sampling and inverse high pass filtering the discrete wavelet coefficients to produce second up sampled and filtered data;

summing the first and second up sampled and filtered data to produce first summed data;

up sampling and inverse low pass filtering the first summed data to produce third up sampled and filtered data;

up sampling and inverse high pass filtering the discrete wavelet coefficients to produce fourth up sampled and filtered data;

summing the third and fourth up sampled and filtered data to produce second summed data;

up sampling and inverse low pass filtering the second summed data to produce fifth up sampled and filtered data;

up sampling and inverse high pass filtering the discrete wavelet coefficients to produce sixth up sampled and filtered data; and

summing the fifth and sixth up sampled and filtered data to produce the compensation data.

6. The vehicle navigation method of claim 5 wherein the performing of an inverse discrete wavelet transform comprises:

performing the inverse discrete wavelet transform by column on the discrete wavelet coefficients to produce a matrix; and

,

performing the inverse discrete wavelet transform by row on the matrix to produce the compensation data.

7. The vehicle navigation method of claim 5 wherein the performing of an inverse discrete wavelet transform comprises:

scaling the discrete wavelet coefficients; and

,

performing the inverse discrete wavelet transform on the scaled coefficients to produce the compensation data.

8. The vehicle navigation method of claim 7 wherein the performing of an inverse discrete wavelet transform comprises:

performing the inverse discrete wavelet transform by column on the scaled coefficients to produce a matrix; and

,

performing the inverse discrete wavelet transform by row on the matrix to produce the compensation data.

9. The vehicle navigation method of claim 8 wherein the compensating of the sensor signal based on the compensation data comprises interpolating the compensation data.

10. The vehicle navigation method of claim 1 wherein the accessing discrete wavelet coefficients comprises performing a lossless decompression of the discrete wavelet coefficients, and wherein the performing of an inverse discrete wavelet transform comprises performing the inverse discrete wavelet transform on the decompressed coefficients to produce the compensation data.

11. The vehicle navigation method of claim 10 wherein the performing of an inverse discrete wavelet transform comprises:

performing the inverse discrete wavelet transform by column on the decompressed coefficients to produce a matrix; and

,

performing the inverse discrete wavelet transform by row on the matrix to produce the compensation data.

12. The vehicle navigation method of claim 10 wherein the performing of an inverse discrete wavelet transform comprises:

scaling the decompressed coefficients; and

,

performing an inverse discrete wavelet transform on the scaled coefficients to produce the compensation data.

13. The vehicle navigation method of claim 12 wherein the performing of an inverse discrete wavelet transform comprises:

performing the inverse discrete wavelet transform by column on the scaled coefficients to produce a matrix; and

,

performing the inverse discrete wavelet transform by row on the matrix to produce the compensation data.

14. The vehicle navigation method of claim 10 wherein the performing of an inverse discrete wavelet transform comprises:

up sampling and inverse low pass filtering the decompressed coefficients to produce first up sampled and filtered data;

up sampling and inverse high pass filtering the decompressed coefficients to produce second up sampled and filtered data;

summing the first and second up sampled and filtered data to produce first summed data;

up sampling and inverse low pass filtering the first summed data to produce third up sampled and filtered data;

up sampling and inverse high pass filtering the decompressed coefficients to produce fourth up sampled and filtered data;

summing the third and fourth up sampled and filtered data to produce second summed data;

up sampling and inverse low pass filtering the second summed data to produce fifth up sampled and filtered data;

up sampling and inverse high pass filtering the decompressed coefficients to produce sixth up sampled and filtered data; and

summing the fifth and sixth up sampled and filtered data to produce the compensation data.

15. The vehicle navigation method of claim 14 wherein the performing of an inverse discrete wavelet transform comprises:

performing the inverse discrete wavelet transform by column on the decompressed coefficients to produce a matrix; and

,

performing the inverse discrete wavelet transform by row on the matrix to produce the compensation data.

16. The vehicle navigation method of claim 14 wherein the performing of an inverse discrete wavelet transform comprises:

scaling the decompressed coefficients; and

,

performing the inverse discrete wavelet transform on the scaled coefficients to produce the compensation data.

17. The vehicle navigation method of claim 16 wherein the performing of an inverse discrete wavelet transform comprises:

performing the inverse discrete wavelet transform by column on the scaled coefficients to produce a matrix; and

,

performing the inverse discrete wavelet transform by row on the matrix to produce the compensation data.

18. The vehicle navigation method of claim 17 wherein the compensating the sensor signal based on the compensation data comprises interpolating the compensation data.

19. A method comprising:

performing a discrete wavelet transform on raw gravitational deflection data to produce coefficients;

performing a lossless compression on the coefficients to produce compressed coefficients; and

,

storing the compressed coefficients in a memory of a vehicle navigation system.

20. The method of claim 19 wherein the discrete wavelet transform is first performed by row until all of the raw gravitational deflection data is so processed and is only then performed by column to produce the coefficients.

21. The method of claim 19 wherein the performing of a lossless compression comprises:

scaling the coefficients; and

,

performing a lossless compression on the scaled coefficients to produce the compressed coefficients.

22. The method of claim 21 wherein the discrete wavelet transform is first performed by row until all of the raw gravitational deflection data is so processed and is only then performed by column to produce the coefficients.

23. The method of claim 21 wherein the performing of a lossless compression on the scaled coefficients comprises:

quantizing the scaled coefficients to produce quantized coefficients; and

,

performing a lossless compression on the quantized coefficients to produce the compressed coefficients.

24. The method of claim 23 wherein the discrete wavelet transform is first performed by row until all of the raw gravitational deflection data is so processed and is only then performed by column to produce the coefficients.

25. The method of claim 23 wherein the performing of a lossless compression on the quantized coefficients comprises:

dividing the quantized coefficients into regions; and

,

performing a lossless compression on the regional quantized coefficients to produce the compressed coefficients.

26. The method of claim 25 wherein the discrete wavelet transform is first performed by row until all of the raw gravitational deflection data is so processed and is only then performed by column to produce the coefficients.

27. The method of claim 19 wherein the performing of a discrete wavelet transform comprises:

high pass filtering and down sampling the raw gravitational deflection data to produce first filtered and down sampled data;

low pass filtering and down sampling the raw gravitational deflection data to produce second filtered and down sampled data;

high pass filtering and down sampling the second filtered and down sampled data to produce third filtered and down sampled data;

low pass filtering and down sampling the second filtered and down sampled data to produce fourth filtered and down sampled data;

high pass filtering and down sampling the fourth filtered and down sampled data to produce fifth filtered and down sampled data; and

,

low pass filtering and down sampling the fourth filtered and down sampled data to produce sixth filtered and down sampled data;

wherein the first, third, fifth, and sixth filtered and down sampled data comprise the coefficients.

28. The method of claim 27 wherein the discrete wavelet transform is first performed by column until all of the raw gravitational deflection data is so processed and is only then performed by row to produce the coefficients.

29. The method of claim 27 wherein the performing of a lossless compression comprises:

scaling the coefficients; and

,

performing a lossless compression on the scaled coefficients to produce the compressed coefficients.

30. The method of claim 29 wherein the discrete wavelet transform is first performed by column until all of the raw gravitational deflection data is so processed and is only then performed by row to produce the coefficients.

31. The method of claim 29 wherein the performing of a lossless compression on the scaled coefficients comprises:

quantizing the scaled coefficients to produce quantized coefficients; and

,

performing a lossless compression on the quantized coefficients to produce the compressed coefficients.

32. The method of claim 31 wherein the discrete wavelet transform is first performed by column until all of the raw gravitational deflection data is so processed and is only then performed by row to produce the coefficients.

33. The method of claim 32 wherein the performing of a lossless compression on the quantized coefficients comprises:

dividing the quantized coefficients into regions; and

,

performing a lossless compression on the regional quantized coefficients to produce the compressed coefficients.

34. The method of claim 33 wherein the discrete wavelet transform is first performed by column until all of the raw gravitational deflection data is so processed and is only then performed by row to produce the coefficients.

35. A method for using compressed wavelet transform coefficients in order to correct outputs of inertial sensors for gravitational deflections, wherein the compressed wavelet transform coefficients comprise gravitational deflection data that is first converted to discrete wavelet coefficients by a discrete wavelet transform and wherein the wavelet transform coefficients are then compressed to form the compressed wavelet transform coefficients, and wherein the method comprises:

generating an inertial sensor signal relating to a navigational state of a vehicle;

accessing the compressed discrete wavelet coefficients from a memory based on a position of the vehicle;

decompressing the accessed coefficients;

performing an inverse discrete wavelet transform on the decompressed coefficients to produce gravitational compensation data; and

,

compensating the inertial sensor signal based on the gravitational compensation data.

36. The vehicle navigation method of claim 35 wherein the wavelet transform coefficients is scaled by a first scaling factor before compression, wherein the method further comprises scaling the decompressed coefficients by a second scaling factor prior to the performing of the inverse discrete wavelet transform, and wherein the second scaling factor is the inverse of the first scaling factor.

37. The vehicle navigation method of claim 35 wherein the performing of an inverse discrete wavelet transform comprises performing the inverse discrete wavelet transform first by column and then by row.

38. The vehicle navigation method of claim 37 wherein the wavelet transform coefficients is scaled by a first scaling factor before compression, wherein the method further comprises scaling the decompressed coefficients by a second scaling factor prior to the performing of the inverse discrete wavelet transform, and wherein the second scaling factor is the inverse of the first scaling factor.

39. The vehicle navigation method of claim 35 wherein the performing of an inverse discrete wavelet transform comprises:

up sampling and inverse low pass filtering the decompressed coefficients to produce first up sampled and filtered data;

up sampling and inverse high pass filtering the decompressed coefficients to produce second up sampled and filtered data;

summing the first and second up sampled and filtered data to produce first summed data;

up sampling and inverse low pass filtering the first summed data to produce third up sampled and filtered data;

up sampling and inverse high pass filtering the decompressed coefficients to produce fourth up sampled and filtered data;

summing the third and fourth up sampled and filtered data to produce second summed data;

up sampling and inverse low pass filtering the second summed data to produce fifth up sampled and filtered data;

up sampling and inverse high pass filtering the decompressed coefficients to produce sixth up sampled and filtered data; and

summing the fifth and sixth up sampled and filtered data to produce the compensation data.

40. The vehicle navigation method of claim 39 wherein the wavelet transform coefficients is scaled by a first scaling factor before compression, wherein the method further comprises scaling the decompressed coefficients by a second scaling factor prior to the performing of the inverse discrete wavelet transform, and wherein the second scaling factor is the inverse of the first scaling factor.

41. The vehicle navigation method of claim 39 wherein the performing of an inverse discrete wavelet transform comprises performing the inverse discrete wavelet transform first by column and then by row.

42. The vehicle navigation method of claim 41 wherein the wavelet transform coefficients is scaled by a first scaling factor before compression, wherein the method further comprises scaling the decompressed coefficients by a second scaling factor prior to the performing of the inverse descrete wavelet transform, and wherein the second scaling factor is the inverse of the first scaling factor.

43. The vehicle navigation method of claim 42 wherein the compensating the sensor signal based on the compensation data comprises interpolating the compensation data.