Scribd
Upload
95 views27 pages

Presented by Harsha Thomas S M.Tech

The document discusses zero-crossings, which refer to points where a signal changes from positive to negative. Zero-crossings can be used to reconstruct bandlimited signals from their discrete samples. Specifically, zero-crossings provide an efficient way to design analog-to-digital converters that only sample the signal at zero-crossing points, potentially reducing power and complexity compared to traditional sampling. The document also discusses using zero-crossing rates for applications like audio classification and spectrum analysis.

Uploaded by

Mahesh Kumar
Copyright
© Attribution Non-Commercial (BY-NC)
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
95 views27 pages

Presented by Harsha Thomas S M.Tech

The document discusses zero-crossings, which refer to points where a signal changes from positive to negative. Zero-crossings can be used to reconstruct bandlimited signals from their discrete samples. Specifically, zero-crossings provide an efficient way to design analog-to-digital converters that only sample the signal at zero-crossing points, potentially reducing power and complexity compared to traditional sampling. The document also discusses using zero-crossing rates for applications like audio classification and spectrum analysis.

Uploaded by

Mahesh Kumar
Copyright
© Attribution Non-Commercial (BY-NC)
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
You are on page 1/ 27

Presented By Harsha Thomas S2 M.

Tech

3/14/2013

Zero-crossing is a commonly used term in electronics, mathematics, sound, and image processing. In mathematical terms, a "zero-crossing" is a point where the sign of a function changes (e.g. from positive to negative), represented by a crossing of the axis (zero value) in the graph of the function. In electronics, the zero-crossing is the instantaneous point at which there is no voltage present. In a sine wave or other simple waveform, this normally occurs twice during each cycle.
3/14/2013 2

Provides an efficient method for reconstructing a bandlimited signal in the discrete domain from its crossings. The method makes it possible to design A/D converters that only deliver the crossings, which are then used to interpolate the input signal at arbitrary instants. Potentially, it may allow for reductions in power consumption and complexity in these converters.

3/14/2013

Sampling is the reduction of a continuous signal to a discrete signal. Sampler produces samples equivalent to the instantaneous value of the continuous signal at the desired points.

3/14/2013

Convolution with pulse creates replicas at pulse location This tells us that the impulse train modulator Creates images of the Fourier transform of the input signal Images are periodic with sampling frequency If s< N sampling maybe irreversible due to aliasing of images

3/14/2013

Transmit the digital samples from one point to another, using digital electronics, rather than analog electronics. Disadvantages: Introduces distortion, because we have limited resolution on the ADC and have limited frequency of sample rate.

3/14/2013

Audio signal classification.


The zero-crossing rate (ZCR) provides a rough estimate of the frequency content of a signal in time-domain at low computational cost. Thus, the ZCR has been widely used in many signal processing applications such as: 1) spectral estimation, 2) voice activity detection, 3) voiced/unvoiced speech discrimination, 4) speech recognition, 5) audio sound classification, fluid mechanics, 6) Doppler frequency estimation, and 7) time-delay estimation

The distribution of ZCRs is used for speech and music discrimination. The number of zero-crossings is strongly related with dominant frequency of a signal.

3/14/2013

Zero crossing based spectrum analyzer:


Spectrum analysers are used for measuring the distribution of signal energy in frequency. zero-based spectrum analyzer determines the DFT coefficients of the signal from the zero axis crossings of the signal rather than from its amplitude samples. The zero-based approach requires no sampling and quantization.

3/14/2013

Representation of time-limited bandpass signal by discrete frequency values. The locations along the frequency axis at which the fourier transform of the signal cross zero level. Sum-of-Sincs model is used. Signal is reconstructed by solving eigen value problem.

3/14/2013

Key idea is to relate fourier coefficients to zero crossings of the fourier transform of the signal. Fourier transform of the signal is computed using filter banks.

3/14/2013

10

3/14/2013

11

X(t) has been obtained by windowing a bandpass signal s(t) using window function w(t). Fourier transform of x(t):

3/14/2013

12

Obtain delayed version of x(t).

3/14/2013

13

The real part of the delayed signal shows more number of axis crossings . Number of additional zero-crossings depends on time delay. Obtain periodic extension of the signal.

Window the periodic extension by a rectangular window whose frequency domain representation contains sinc functions.
3/14/2013 14

3/14/2013

15

Determine zero crossings. Determine fourier coefficients using eigen value problem. Signal can be reconstructed using finite Fourier series.

3/14/2013

16

3/14/2013

17

The analog-to-discrete (A/D) conversion is the first step for the discrete-time processing of continuous signals. This conversion is fundamentally based on the Sampling Theorem. Encoding using zero-crossings provide a way to convert an analog signal into discrete values using SOS model.

3/14/2013

18

1)

R. Kumaresan , Encoding bandpass signals using zero/level crossings : A model based approach, IEEE Trans. Speech Audio Process., vol. 18, no. 1, jan 2010.

2)

R. Kumaresan and Y. Wang, On the relationship between line spectrum pairs and zero-crossings of band-pass signals, IEEE Trans. Speech Audio Process., vol. 9, no. 4, pp. 458461, May 2001.
S. M. Kay and R. Sudhaker, A zero crossing-based spectrum analyzer, IEEE Trans. Acoust., Speech, Signal Process., vol. ASSP-34, no. 1, pp. 96104, Jan. 1986. D. Kim, S. Lee, and R. Kil, Auditory processing of speech signals for robust speech recognition in real-world noisy environments, IEEE Trans. Speech Audio Process., vol. 7, no. 1, pp. 5569, Jan. 1999.
3/14/2013 19

3)

4)

THANK YOU

3/14/2013

20

Fourier series expansion for

where and are Fourier coefficients. The integers M and N are such that M and N is bandlimited to

3/14/2013

21

Window

by a rectangular window

We get The model for

is

3/14/2013

22

where is the Fourier transform of convolution operation. is given by:

and * denote

where and denote the real and imaginary parts of respectively and they are frequency domain sinc functions.
3/14/2013 23

The real and imaginary parts of and , respectively.

is denoted as

Since and are expressed as a linear combination of shifted versions of the Sinc functions, the model is called SOS model.
3/14/2013 24

are the locations along the frequency axis at which the real part is zero. , k = 1,2,,q are the locations corresponding to
k = 1,2,.,p

Writing in matrix-vector notation

3/14/2013

25

c contains unknown Fourier coefficients.

The submatrices and have p and q rows, respectively. c is a unique vector in the null space of X. Estimate the vector of Fourier coefficients c by minimizing the quadratic form: .

3/14/2013

26

The solution vector c is the eigenvector of XTX corresponding to its smallest eigenvalue. Fourier transform of delayed signal Fourier coefficients reconstruct the signal using Fourier series.

3/14/2013

27

You might also like