Systems And Methods For Estimating A Wavelet Transform With A Goertzel Technique

1 . A method for processing a signal by estimating a wavelet transform of the signal using at least one scale, the at least one scale being associated with at least one data sample, the method comprising:

generating, using electronic processing equipment, a magnitude value for each sample in each of the at least one scale based at least in part on a Goertzel technique, wherein each magnitude value estimates a wavelet transform; and

storing, in an electronic storage device, the generated magnitude values.

2 . The method of claim 1 , wherein the generating comprises:

initializing a predetermined coefficient value and temporary storage values for each sample in each of the at least one scale;

for each of the at least one data sample in each of the at least one scale:

computing a recursion value based on a first function of the temporary storage values corresponding to the at least one data sample in each of the at least one scale, the predetermined coefficient value, and the corresponding at least one data sample in each of the at least one scale; and

updating the temporary storage values corresponding to the at least one data sample in each of the at least one scale based on the computed recursion value;

repeating the computing and the updating a predetermined number of times; and

computing the magnitude value for each of the at least one data samples in each of the at least one scale based on a second function of (1) the predetermined coefficient value and (2) the updated temporary storage values corresponding to the at least one data sample in each of the at least one scale.

3 . The method of claim 2 , wherein:

an output of the first function is computed in accordance with:

coeff*Q1−Q2+sample;

where coeff is the predetermined coefficient value, Q1 and Q2 are first and second of the temporary values and sample is the corresponding at least one sample in each of the at least one scale; and

an output of the second function is computed in accordance with:

Q1*Q1−Q2*Q2−Q1*Q2*coeff.

4 . The method of claim 3 , wherein:

the predetermined coefficient value is computed as a multiple of the cosine function of a value corresponding the predetermined number of times, a sample rate of the at least one scale, and a target frequency value for the at least one scale.

5 . The method of claim 2 wherein the predetermined number of times corresponds to a number of scales in the wavelet transform.

6 . The method of claim 1 further comprising providing the magnitude values of each of the at least one data sample in each of the at least one scale as the estimated wavelet transform scalogram.

7 . The method of claim 1 , wherein the magnitude values for each of the at least one data sample in each of the at least one scale are computed in parallel.

8 . The method of claim 1 , wherein the magnitude values are computed without convolution operations.

9 . The method of claim 1 further comprising applying a window to the at least one data sample in each of the at least one scale to change the filter response of the at least one scale.

10 . The method of claim 9 , wherein the window is a Gaussian window or a Hamming window.

11 . A system for processing a signal by estimating a wavelet transform of the signal using at least one scale, the at least one scale being associated with at least one data sample, the system comprising:

an electronic storage device; and

processing circuitry configured to:

generate, using electronic processing equipment, a magnitude value for each sample in each of the at least one scale based at least in part on a Goertzel technique, wherein each magnitude value estimates a wavelet transform; and

store, in the electronic storage device, the generated magnitude values.

12 . The system of claim 11 , wherein the processing circuitry is further configured to:

initialize a predetermined coefficient value and temporary storage values for each sample in each of the at least one scale;

for each of the at least one data sample in each of the at least one scale:

compute a recursion value based on a first function of the temporary storage values corresponding to the at least one data sample in each of the at least one scale, the predetermined coefficient value, and the corresponding at least one data sample in each of the at least one scale; and

update the temporary storage values corresponding to the at least one data sample in each of the at least one scale based on the computed recursion value;

repeat the computing and the updating a predetermined number of times; and

compute the magnitude value for each of the at least one data samples in each of the at least one scale based on a second function of (1) the predetermined coefficient value and (2) the updated temporary storage values corresponding to the at least one data sample in each of the at least one scale.

13 . The system of claim 12 , wherein:

an output of the first function is computed in accordance with:

coeff*Q1−Q2+sample;

where coeff is the predetermined coefficient value, Q1 and Q2 are first and second of the temporary values and sample is the corresponding at least one sample in each of the at least one scale; and

an output of the second function is computed in accordance with:

Q1*Q1+Q2*Q2−Q1*Q2*coeff.

14 . The system of claim 13 , wherein:

the predetermined coefficient value is computed as a multiple of the cosine function of a value corresponding the predetermined number of times, a sample rate of the at least one scale, and a target frequency value for the at least one scale.

15 . The system of claim 12 wherein the predetermined number of times corresponds to a number of scales in the wavelet transform.

16 . The system of claim 11 wherein the processing circuitry is further configured to provide the magnitude values of each of the at least one data sample in each of the at least one scale as the estimated wavelet transform scalogram.

17 . The system of claim 11 , wherein the magnitude values for each of the at least one data sample in each of the at least one scale are computed in parallel.

18 . The system of claim 11 , wherein the magnitude values are computed without convolution operations.

19 . The system of claim 11 further comprising applying a window to the at least one data sample in each of the at least one scale to change the filter response of the at least one scale.

20 . The system of claim 19 , wherein the window is a Gaussian window or a Hamming window.