Abstract

In this paper, we propose the use of geometric moments to the field of nonblind image deblurring. Using the developed relationship of geometric moments for original and blurred images, a mathematical formulation based on the Euler–Lagrange identity and variational techniques is proposed. It uses an iterative procedure to deblur the image in moment domain. The theoretical framework is validated by a set of experiments. A comparative analysis of the results obtained using the spatial and moment domains are evaluated using a quality assessment method known as the Blind/Reference-less Image Spatial Quality Evaluator (BRISQUE). The results show that the proposed method yields a higher quality score when compared with the spatial domain method for the same number of iterations.

© 2014 Optical Society of America

Tchebichef moment based restoration of Gaussian blurred images

Ahlad Kumar, Raveendran Paramesran, Chern-Loon Lim, and Sarat C. Dass
Appl. Opt. 55(32) 9006-9016 (2016)

Adaptive deblurring of noisy images

S. Susan Young, Ronald G. Driggers, Brian P. Teaney, and Eddie L. Jacobs
Appl. Opt. 46(5) 744-752 (2007)

Nonlocal means image denoising using orthogonal moments

Ahlad Kumar
Appl. Opt. 54(27) 8156-8165 (2015)

Dictionary learning approach for image deconvolution with variance estimation

Hang Yang, Ming Zhu, Xiaotian Wu, Zhongbo Zhang, and Heyan Huang
Appl. Opt. 53(25) 5677-5684 (2014)

Blind deconvolution of fluorescence micrographs by maximum-likelihood estimation

Vijaykumar Krishnamurthi, Yi-Hwa Liu, Santosh Bhattacharyya, James N. Turner, and Timothy J. Holmes
Appl. Opt. 34(29) 6633-6647 (1995)