Efficient Multi Focus Image Fusion Technique Optimized Using MOPSO for Surveillance Applications

2. Related Works

Xia Xiaohua et al. (2014) have used a global mapping model for initial registration as it one of the essential pre-processing step of image fusion. Focus measure operator was used for searching the best focused pixels in the sequence of images. Probability filtering is used for enhancing the robustness of focus measure and an accurate registration is performed for extracting the common features of the focused regions. Even though this method yields better results, it has high complexity due to repeated registration. Chandrakanth et al. (2014) have discussed an image fusion system for multi sensor and multiband remote sensing data. This image fusion system focuses on selection of data, pre-processing, registration and fusion for multi sensor and multiband images. A robust image registration method and appropriate fusion techniques are analyzed. The discussed technique yields better performance when compared to well-known techniques. Bai Xiangzhi et al. (2015) have presented a quad tree based decomposition methodology. In this strategy, the input images are decomposed into blocks with various sizes and focused regions are detected by using Sum of the Weighted Modified Laplacian (SWML). Akinlar Mehmet Ali et al. (2013) have analyzed a hybrid method for deformable matching of magnetic resonance (MR) images using the benefits of both wavelet and variational calculus. This method is implemented based on the Gabor wavelet energy maps of MR images. Experimental results proved that this method yields better results compared to the variation-based technique and wavelet based technique.