An Automatic Calculation Method of MTF and the Application in Blurred Images Restoration

Abstract:

Modulation transfer function (MTF) is a comprehensive index for the objective reflection of the quality of the imaging systems. In the field of image processing, the slanted-edge method is usually adopted to compute MTF for images. However, the sub-images with slanted edges are extracted from original image by subjective judgment and manually operation, which is bound to lead to inefficiency of calculation, low accuracy and instability of results with the participation of humans. Aiming at the above problem, this paper presents an automatic MTF calculation method for blurred images and applies it to the process of image restoration by developing a two-dimensional MTF filter and utilizing it into conventional restoration methods such as the Wiener filtering, least squares filtering and Lucy-Richardson algorithm. Experiment results indicate the proposed method achieved an automatic, fast and accurate MTF computation for blurred images, and MTF-based restoration methods were superior to traditional ones in restoration effects.

• Periodical: Applied Mechanics and Materials (Volume 731)
• Pages: 141-146
• DOI: https://doi.org/10.4028/www.scientific.net/AMM.731.141
• Online since: January 2015
• Authors: Yuan Yuan*, Yao Hua Yi, Min Jing Miao
• Keywords: Hough Transform, Image Restoration, MTF Filter, Slanted-Edge Method

References

[1] Wang Hongnan, Zhong Wen, Wang Jing and Xia Desheng: Journal of Image and Graphics, Vol. 9 (2004), pp.64-67.

[2] Li Tiecheng, Tao Xiaoping and Feng Huajun: Acta Optica Sinica, 30 (2010), pp.2891-2897.

[3] McGillem C D, Anuta P E and Malaret E: LARS Technical Reports, 1983, p.81.

[4] Leger Dominique, Duffaut Joel and Robinet Francoise: International Geoscience and Remote Sensing Symposium (IGARSS), 4(1994), p.2010-(2012).

DOI: 10.1109/igarss.1994.399638

[5] Xinhua Cao, H. K. Huang and Shyh Linaglou: International Society for Optics and Photonics, 3922(2000): pp.580-589.

[6] Yao Haigen: Publishing and Printing, 4 (2009), pp.2-7.

[7] ISO 20462: Photography-Psychophysical experimental methods for estimating image quality (2005).