Multi-focus image fusion based on probability filtering and region correction
| Autor: | Zibing Yang, Xiaohua Xia, Yunshi Yao, Shida Wu, Lijuan Yin, Haochen Li |
|---|---|
| Rok vydání: | 2018 |
| Předmět: |
Image fusion
Pixel Physics::Instrumentation and Detectors business.industry Computer science Multi focus ComputingMethodologies_IMAGEPROCESSINGANDCOMPUTERVISION 020206 networking & telecommunications Pattern recognition 02 engineering and technology Filter (signal processing) Image (mathematics) Control and Systems Engineering Computer Science::Computer Vision and Pattern Recognition Signal Processing 0202 electrical engineering electronic engineering information engineering 020201 artificial intelligence & image processing Computer Vision and Pattern Recognition Artificial intelligence Depth of field Electrical and Electronic Engineering business Focus (optics) Software |
| Zdroj: | Signal Processing. 153:71-82 |
| ISSN: | 0165-1684 |
| Popis: | Multi-focus image fusion is a common approach for extending the limited depth of field of cameras. However, existing focus measures cannot accurately identify focused pixels in multi-focus image fusion. It leads to the loss of focused pixels and the introduction of defocused pixels in the fused image. To solve this problem, a novel pixel-based fusion algorithm is proposed. In this algorithm, misidentified pixels are classified into two types: the dispersed and the clustered. Probability filtering and region correction are proposed to correct the dispersed and the clustered misidentified pixels, respectively. Experimental results show that the misidentified pixels are corrected by probability filtering and region correction, and the fusion algorithm is superior to the conventional algorithms and the state-of-the-art algorithms in terms of both subjective evaluation and objective evaluation. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|
Full Text from ScienceDirect