I trying to use the entropy and joint entropy for base wavelet selection but faced me problem
i considered the original signal X and its wavelet coefficients Y as two variables. but the length of wavelet coefficients Y is not equal to the length of the original signal X. the joint entropy required that the two variable X, Y must equal in length. pl how can I let the wavelet coefficients length equal to original signal length
Dear Ahmed,
There are works suggesting use of joint entropy for mother wavelet selection. I attached one for you to read. Best of luck,
Dear Ahmed
you want to calculate joint entropy between signal and its wavelet coefficient. The number of coefficients should be same but DWT decomposition reduced number by 1/2 in next level. you can use SWT decomposition (MATLAB Function) to get same number of coefficients.
Hi Ahmed, you have a digital signal X, wavelet is Y. The DECOMPOSITION of your signal X on the waveleght is what you are after. I am not sure if you understand that, but as K.K. Gupta suggested, use SWT but make sure you understand what you are doing. Practically when you use a wavelet, you are calculating how many "pieces" of the siganl are similar to the Mother wavelet at that level. If you have 2 "matching pieces" that will give you 2. The original idea is from the cosine decomposition of any signal/Fourier transform.i.e. the jus instead the small cosine function you are using a wavelet.
- Badges
- Science topic
- Similar topics
- Filing