Engineering Commons^™

A Defect Detection Algorithm For Sequential And Parallel Computers, Ralf Hross

Theses

The comparison of images containing a single object of interest, where one of them contains a model object, is frequently used for defect detection/identification. This is often a problem of interest to industrial applications. To make this task more efficient, the implementation on parallel computers should be a major objective. This thesis introduces sequential and parallel versions of an algorithm that compares original(reference) and processed images in the time and frequency domains. The emphasis is on parallel implementation. This comparison may help to detect changes in images. Extracted data is also compared to database data in an attempt to pinpoint …

Motion Compensation Using Correlation-Feedback, Zebiba Shifa

Theses

Motion compensation is widely used for exploiting temporal redundancies in the coding of image sequences. Accurate estimation of motion information in image sequences is important in motion-compensated coding. Different approaches have been used to estimate motion to obtain the motion-compensated frame difference signal.

This work uses the correlation-feedback approach to estimate the velocity or the optic flow of the moving image pixel. After the motion of the pixel is estimated. the motion-compensated frame difference signal is found by subtracting the current. frame from the predicted frame.

This correlation-feedback approach estimates the true motion vector of moving image accurately. Consequently, the …

A Hybrid Low Bit-Rate Video Codec Using Subbands And Statistical Modeling, Ferhat Cakrak

Theses

A hybrid low bit-rate video codes using subbands and statistical modeling is proposed in this thesis. The redundancy within adjacent video frames is exploited by motion estimation and compensation. The Motion Compensated Frame Difference (MCFD) signals are decomposed into 7 subbands using 2-D dyadic tree structure and separable filters. Some of the subband signals are statistically modeled by using the 2-D AR(1) technique. The model parameters provide a representation of these subbands at the receiver side with a. certain level of error. The remaining subbands are compressed employing a classical waveform coding technique, namely vector quantization (VQ).

It is shown …

A Generalized, Parametric Pr-Qmf/Wavelet Transform Design Approach For Multiresolution Signal Decomposition, Hakan Caglar

Dissertations

This dissertation aims to emphasize the interrelations and the linkages of the theories of discrete-time filter banks and wavelet transforms. It is shown that the Binomial-QMF banks are identical to the interscale coefficients or filters of the compactly supported orthonormal wavelet transform bases proposed by Daubechies.

A generalized, parametric, smooth 2-band PR-QMF design approach based on Bernstein polynomial approximation is developed. It is found that the most regular compact support orthonormal wavelet filters, coiflet filters are only the special cases of the proposed filter bank design technique.

A new objective performance measure called Non-aliasing Energy Ratio(NER) is developed. Its merits …