小波转换应用於二维碎形布朗运动估测.pptVIP

小波轉換應用於二維碎形布朗運動估測,抗雜訊之影像壓縮,及影像序列上物件追蹤Estimation of 2D Fractional Brownian Motion, Robust Image Compression, and Object Tracking in Image Sequences using Wavelets 研究生:劉震昌 指導教授:陳銘憲博士 黃文良博士 研究方向 Wavelet-based image processing Estimation of 2D fractional Brownian motion Error-resilient image compression and transmission over IP/wireless network Object tracking in image sequences using active contour model 論文著作 期刊論文 “Estimation of 2-D noisy fractional Brownian motion and its applications using wavelets”, IEEE trans. on Image Processing, 2000. “A block-based channel-optimized source coding for EZW-like progressive codec using rate-distortion analysis”, in revise, IEEE trans. on Image Processing. 論文著作(2) 會議論文 “Parameter estimation and denoising of 2-D noisy fractional Brownian motion using non-orthogonal wavelet”, IEEE Symposium on Time-Frequency and Time-Scale Analysis, 1998. “Robust block-based EZW image compression with channel noise optimized rate-distortion functions”, IEEE ICIP, 1999. “A multi-channel channel-optimized scheme for EZW using rate-distortion functions”, IEEE ICIP, 2000. “Wavelet-based active contour model for object tracking”, IEEE ICIP, 2001. Review of Wavelet Transform What are the wavelets? Wavelets are a family of functions produced from a “mother wavelet” by translation and dilation: Admissibility condition Discrete Wavelet Transform Wavelet decomposition (Mallat’89 PAMI) Discrete Wavelet Transform: Implementation Filterbank implementation Discrete Wavelet Transform: Mallat’s Wavelet (PAMI’92) Let q(x) be a smoothing function x-direction mother wavelet: y-direction mother wavelet: Suitable for edge detection Estimation of 2-D Fractional Brownian Motion(fBm) and its Applications using Wavelets Outline Introduction to fBm fBm parameter estimation fBm de-noising Applications 2-D fBm: examples 2-D Isotropic Fractional Brownian Motion Brownian motion: real random function with independent Gaussian increments B(t2)-B(t1) has