This text introduces and investigates estimation algorithms for signal registration. Signal registration can be defined as the problem of estimating a varying displacement between two random processes. This problem has great importance in many applications, perhaps the most familiar of which lies in the field of passive sonar. The bearing of the signal source is related to the time delay (displacement) between two random waveforms. In the general signal registration problem, the time delay is itself a random function of time. Part one of this book is devoted to the problem of variable time delay estimation. Another application of the signal registration lies in image processing where registration is called "motion estimation". By determining the relative displacement (motion) between the intensities of consecutive image frames, one can encode an image more efficiently. In this application, the displacement is a position-dependent vector process. Part two discusses the problem of nonuniform image motion estimation.

Table of Contents:
Acknowledgements          ix 1  Introduction to Signal Registration          1   PART I     VARIABLE TIME DELAY ESTIMATION          3 2  Introduction to Time Delay Estimation          5 3  A New Recursive Estimator          9  4  Delay Estimation based on the MAP Criterion          13     1. Implementation using the MAP Estimation          13     2. Estimation Analysis          16     3. Simulation Experiments          19 5. Delay Estimation based on the ML Criterion          46     1. The Generalized Maximum Likelihood Algorithm (GML)          46     2. Implementation using the GML Estimator          48     3. Simulation Experiments          50 6  Delay Estimation Based on the MMSE Criterion          60     1. The Minimum Mean Squared Error (MMSE1) Algorithm          60      2. Implementation using the MMSE1 Algorithm          62     3. Simulation Experiments using the MMSE1 Algorithm          62     4. The Minimum Mean Squared Error (MMSE2) Algorithm          65     5. Simulation Experiments using the MMSE2 Algorithm          71 7  Concluding Remarks on Part 1          84   PART II     NONUNIFORM IMAGE MOTION ESTIMATION          87 8  Introduction to Image Motion Estimation          89 9  Motion Estimation based on the ML Criterion          92     1. The Generalized Maximum Likelihood Algorithm in Two-Dimensional Space          92     2. Implementation Using the GML Algorithm          95     3. Convergence of the GML Algorithm          98     4. Simulation Experiments          103 10  Motion Estimation Based on the MAP Criterion          112       1. The Maximum a Posteriori Algorithm          112       2. Implementation Using the MAP Algorithm          115       3. Determination of [A greek symbol]1 and [B greek symbol]1 in the Motion Estimator (10.20)          117       4. Simulation Experiments          120 11  Motion Estimation in Transformed Domains          125       1. The Transformed Domain Maximum Likelihood (TDML) Algorithm          125       2. Implementation using the TDML Algorithm          126       3. Simulation Experiments          128 12  Estimation of the Motion Coefficients Using Kalman Filtering          132        1. Motion Estimation with Kalman Filtering          132       2. Identification of the Delectable Moving Pixels from Noisy Data          133       3. Representation of the Reduced-Dimension Image Motion Coefficients in State Space          137        4. Generation of the Innovations Process          139       5. Implementation of the Kalman Filter          141        6. Results          143 13 Concluding Remarks on Part II          149   Appendix A.     An Adaptive Realization of the Generalized Maximum Likelihood Algorithm          151 References          157 Author Index          161 Subject Index          163