Multi-focus Image Fusion via Region Mosaicing on Contrast Pyramids
This paper proposes a new approach called region mosaicing on contrast pyramids for multi-focus image fusion. A density-based region growing is developed to construct a focused region mask for multi-focus images. The segmented focused region mask is decom
- PDF / 2,804,020 Bytes
- 11 Pages / 439.37 x 666.142 pts Page_size
- 71 Downloads / 203 Views
Harbin Engineering University, Harbin 150001, China {zhangliguo,sunjianguo,linjunyu}@hrbeu.edu.cn 2 Institute of Information Engineering, Chinese Academy of Sciences, Beijing 100093, China 3 Montana State University, Bozeman 59717, USA [email protected]
Abstract. This paper proposes a new approach called region mosaicing on contrast pyramids for multi-focus image fusion. A density-based region growing is developed to construct a focused region mask for multi-focus images. The segmented focused region mask is decomposed into a mask pyramid, which is then used for supervised region mosaicing on a contrast pyramid. In this way, the focus measurement and the continuity of focused regions are incorporated and the pixel level pyramid fusion is improved at the region level. Objective and subjective experiments show that the proposed region mosaicing on contrast pyramids approach is more robust to noise and can fully preserves the focus information of the multi-focus images, reducing distortions of the fused images. Keywords: Multi-focus image pyramid
Image fusion
Region Mosaic
Contrast
1 Introduction High magnification optical lens, such as microscopes, has a very small depth of field. When capturing an object/scene of depth with optical lens of high magnification typically, only a fraction of the object/scene is in focus. Multi-focus image fusion is a process in which registered images with different focus settings are fused to synthesize an “all-in-focus” image with extended depth of field [1–3]. It plays important roles in microscope imaging [4], optical image de-blurring, shape from focus [4, 5] and image based forensics [6]. Pyramid based approaches have been extensively investigated in the image fusion domain. Some examples include The discrete wavelet transform (DWT) [7, 8], the gradient pyramid [9], the contrast pyramid [10], the Laplacian pyramid [11], the ratio-of-low pass pyramid [12], the shift-invariant DWT [13] and the contour let transform [14–16]. Despite the advantages of pyramid based approaches, they are pixel based and then generally sensitive to noise. Noise pixels often have high contrast, and may be falsely detected as in-focus pixels. Because the fused image obtained by transform domain-based algorithms consider global information, a small change in a © Springer International Publishing Switzerland 2016 Q. Yang et al. (Eds.): WASA 2016, LNCS 9798, pp. 80–90, 2016. DOI: 10.1007/978-3-319-42836-9_8
Multi-focus Image Fusion via Region Mosaicing
81
single coefficient of the fused image in the transformed domain may cause all the pixel values to change in spatial domain [17]. Consequently, distortion artifacts and the loss of contrast information are often observed in fused images. Gradient map filtering [8] and multiple coefficient selection principles [18] are proposed to solve this problem, although they need more parameter fine turning to obtain high objective quality. Compared with pyramid based methods, weighted linear fusion is the most intuitive approach to image fusion [19–21].
Data Loading...
