On convergence of kernel estimators of density with variable window width by dependent observations

Focus measure computes sharpness or high frequency contents in an image. It plays an important role in many image processing and computer vision applications such as shape from focus (SFF) techniques and multi-focus image fusion algorithms. In this chapter, we discuss different focus measures in spatial as well as in the transform domains. In addition, we suggest a novel focus measure in S-transform domain, which is based on the energy of high frequency components. A localized spectrum, by using variable window size, provides a more accurate method of measuring image sharpness as compared to other focus measures proposed in spectral domains. An optimal focus measure is obtained by selecting an appropriate frequency dependent window width. The performance of the proposed focus measure is compared with those of existing focus measures in terms of three dimensional shape recovery and all-in-focus image generation. Experimental results demonstrate the effectiveness of the proposed focus measure.

Download Full-text

Variable window width kernel estimates of probability densities

Probability Theory and Related Fields ◽

10.1007/bf01194494 ◽

1992 ◽

Vol 91 (1) ◽

pp. 133-133

Author(s):

Peter Hall ◽

J. S. Marron

Keyword(s):

Kernel Estimates ◽

Window Width ◽

Probability Densities ◽

Variable Window

Download Full-text

Image Focus Measure Based on Energy of High Frequency Components in S-Transform

Depth Map and 3D Imaging Applications ◽

10.4018/978-1-61350-326-3.ch010 ◽

2011 ◽

pp. 189-208

Author(s):

Muhammad Tariq Mahmood ◽

Tae-Sun Choi

Keyword(s):

High Frequency ◽

Window Size ◽

Accurate Method ◽

Image Sharpness ◽

Focus Measure ◽

Window Width ◽

S Transform ◽

Variable Window ◽

Focus Image ◽

Frequency Components

Focus measure computes sharpness or high frequency contents in an image. It plays an important role in many image processing and computer vision applications such as shape from focus (SFF) techniques and multi-focus image fusion algorithms. In this chapter, we discuss different focus measures in spatial as well as in the transform domains. In addition, we suggest a novel focus measure in S-transform domain, which is based on the energy of high frequency components. A localized spectrum, by using variable window size, provides a more accurate method of measuring image sharpness as compared to other focus measures proposed in spectral domains. An optimal focus measure is obtained by selecting an appropriate frequency dependent window width. The performance of the proposed focus measure is compared with those of existing focus measures in terms of three dimensional shape recovery and all-in-focus image generation. Experimental results demonstrate the effectiveness of the proposed focus measure.

Download Full-text