Separating desired image and signal invariant components from extraneous variations

Williams, William J.; Zalubas, Eugene J.; Hero, Alfred O.

doi:10.1117/12.255439

Cited by 9 publications

(10 citation statements)

References 9 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Using a direct expansion of the scale transform, which has some advantages over the previous discrete scale transform reported by our group [15], a new discrete approximation is obtained which avoids the problem of interpolating and exponentially resampling the data. Let t"eV, dt"eVdx.…”

Section: Appendix: the Direct Scale Transformmentioning

confidence: 99%

See 1 more Smart Citation

Helicopter Transmission Fault Detection via Time-Frequency, Scale and Spectral Methods

Williams¹,

Zalubas²

2000

Mechanical Systems and Signal Processing

View full text Add to dashboard Cite

Section: Appendix: the Direct Scale Transformmentioning

confidence: 99%

“…We have developed discrete forms of the scale transform [15,16] which can be computed e$ciently. One might question the use of the scale transform rather than the more well-known Mellin transform.…”

Section: The Scale Transformmentioning

confidence: 99%

Helicopter Transmission Fault Detection via Time-Frequency, Scale and Spectral Methods

Williams¹,

Zalubas²

2000

Mechanical Systems and Signal Processing

View full text Add to dashboard Cite

“…We have developed discrete forms of the scale transform [14,16] which can be computed efficiently. One might question the use of the scale transform rather than the more wellknown Mellin transform.…”

Section: The Scale Transformmentioning

confidence: 99%

“…One can now readily design TFDs which represent the joint energy of a signal as a function of time and frequency or space-frequency distributions which represent the joint energy of images as space-spatial frequency distributions (two spatial variables x and y and two spatial frequency variables X and Y ). Furthermore, with careful design, these joint distributions can exhibit proper covariances with time, frequency or spatial shifts such that the representation shifts in accordance with these shifts but does not change in its configuration [14]. The well-known spectrogram has been extensively used in speech analysis and it has these useful properties.…”

Section: Marine Mammal Soundsmentioning

confidence: 99%

Untitled

Williams

Zalubas

Nickel

et al. 1998

Multidimensional Systems and Signal Processing

Self Cite

View full text Add to dashboard Cite

Time-frequency (t-f) analysis has clearly reached a certain maturity. One can now often provide striking visual representations of the joint time-frequency energy representation of signals. However, it has been difficult to take advantage of this rich source of information concerning the signal, especially for multidimensional signals. Properly constructed time-frequency distributions enjoy many desirable properties. Attempts to incorporate t-f analysis results into pattern recognition schemes have not been notably successful to date. Aided by Cohen's scale transform one may construct representations from the t-f results which are highly useful in pattern classification. Such methods can produce two dimensional representations which are invariant to time-shift, frequency-shift and scale changes. In addition, two dimensional objects such as images can be represented in a like manner in a four dimensional form. Even so, remaining extraneous variations often defeat the pattern classification approach. This paper presents a method based on noise subspace concepts. The noise subspace enhancement allows one to separate the desired invariant forms from extraneous variations, yielding much improved classification results. Examples from sound classification are discussed.

show abstract

“…Our technique for pattern classication uses STIR images decomposed into an orthonormal set of descriptors, using a concept borrowed from Pisarenko's harmonic decomposition [5,8]. The Karhonen-Lo eve transform is a means of accomplishing this.…”

Section: Classication Of Patternsmentioning

confidence: 99%

Shift and scale invariant detection

Zalubas

O'Neill²,

Williams

et al.

1997 IEEE International Conference on Acoustics, Speech, and Signal Processing

View full text Add to dashboard Cite

Dierent signal realizations generated from a given source may not appear the same. Time shifts, frequency shifts, and scales are among the signal variations commonly encountered. Time-frequency distributions (TFDs) covariant to time and frequency shifts and scale changes reect these variations in a predictable manner. Based on such TFDs, representations invariant to these signal distortions are possible. Presented here are two approaches for discriminating between signal classes where within class translation and scale variation occur. The rst method uses an auto-correlation followed by a scale transform to achieve the invariances. The second method treats the TFD as a twodimensional probability density function and applies a transformation that removes the mean and variance to provide the shift and scale invariance. Each method employs discrimination mechanisms to yield powerful results.

show abstract

Separating desired image and signal invariant components from extraneous variations

Cited by 9 publications

References 9 publications

Helicopter Transmission Fault Detection via Time-Frequency, Scale and Spectral Methods

Helicopter Transmission Fault Detection via Time-Frequency, Scale and Spectral Methods

Untitled

Shift and scale invariant detection

Contact Info

Product

Resources

About