Kernel Regression for Image Processing and Reconstruction

• Published in: IEEE transactions on image processing Vol. 16; no. 2; pp. 349 - 366
• Main Authors: Takeda, H., Farsiu, S., Milanfar, P.
• Format: Journal Article
• Language: English
• Published: New York, NY IEEE 01.02.2007; Institute of Electrical and Electronics Engineers; The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
• Subjects: Algorithms; Applied sciences; Artificial Intelligence; Bilateral filter; Charge coupled devices; Contact; Costs; denoising; Detection, estimation, filtering, equalization, prediction; Digital images; Exact sciences and technology; Filters; fusion; Image Enhancement - methods; Image Interpretation, Computer-Assisted - methods; Image processing; Image reconstruction; Information Storage and Retrieval - methods; Information, signal and communications theory; Interpolation; irregularly sampled data; Kernel; kernel function; kernel regression; Kernels; local polynomial; Miscellaneous; Noise reduction; nonlinear filter; nonparametric; Reconstruction; Regression; Regression Analysis; scaling; Signal and communications theory; Signal processing; Signal Processing, Computer-Assisted; Signal, noise; Spatial resolution; spatially adaptive; super-resolution; Telecommunications and information theory; scaling; Image processing; irregularly sampled data; spatially adaptive; Noise reduction; Superresolution; nonlinear filter; super-resolution; Non linear filter; local polynomial; Algorithm; nonparametric; Image reconstruction; Kernel method; denoising; fusion; Kernel function; Interpolation; Bilateral filter; Signal processing; kernel regression; Algorithm analysis; Image denoising
• ISSN: 1057-7149; 1941-0042
• DOI: 10.1109/TIP.2006.888330

In this paper, we make contact with the field of nonparametric statistics and present a development and generalization of tools and results for use in image processing and reconstruction. In particular, we adapt and expand kernel regression ideas for use in image denoising, upscaling, interpolation, fusion, and more. Furthermore, we establish key relationships with some popular existing methods and show how several of these algorithms, including the recently popularized bilateral filter, are special cases of the proposed framework. The resulting algorithms and analyses are amply illustrated with practical examples