New Fundamental Formulas of Image Restoration in Spatial and Frequency Domains

2018/11/03 by Huang, Changcun
#FOS: Electrical engineering #Image and Video Processing (eess.IV) #electronic engineering #information engineering

paper · doi:10.48550/arxiv.1811.02423

Abstract

Circular convolutions and the corresponding frequency domain formula are fundamentally important in image restoration; however, in this paper, we'll prove that the usual computing method of circular convolutions violates the physical meaning of blur producing. Especially for the image restoration algorithms in frequency domain, this violation will affect the restoration result. Relevant problems are proved rigorously and modified formulas are given in both spatial and frequency domains. Experiments are done to show the effects of new formulas. For clarity of proving, the one-dimensional case is dealt with first, which may be useful in one-dimensional signal processing.

Related

• Simple method to eliminate blur based on Lane and Bates algorithm
• On the Logic Elements Associated with Round-Off Errors and Gaussian Blur in Image Registration: A Simple Case of Commingling
• Blind Image Deblurring by Spectral Properties of Convolution Operators
• Spatial Frequency Loss for Learning Convolutional Autoencoders
• Out-of-focus Blur: Image De-blurring