Conclusion
Conclusion¶
To what extent is the guided filter fusion applicable to real world use case ?
For HDR, most cameras are already capable to take at once several photographs with different exposition parameters. Granted we are able to tune hyperparameters adaptatively (which still is an open problem), the method could be used for fusion. For multi-focus fusion however, we saw how challenging it was to obtain good photographs. First, few cameras (none except lytro ?) are able to take at once several photographs with different focal lengths. Second, it requires a preprocessing registration step that is not obvious to design systematically.
In both cases anyway, the computations are too slow to be used in real-time applications (for instance as a smartphone app), or at least our implementation is. For reference, it takes around 10 seconds to do the fusion of our 8 pen images.
References
- HST10
Kaiming He, Jian Sun, and Xiaoou Tang. Guided image filtering. In Kostas Daniilidis, Petros Maragos, and Nikos Paragios, editors, Computer Vision – ECCV 2010, 1–14. Berlin, Heidelberg, 2010. Springer Berlin Heidelberg.
- LKH13
Shutao Li, Xudong Kang, and Jianwen Hu. Image fusion with guided filtering. IEEE Transactions on Image Processing, 22(7):2864–2875, 2013. doi:10.1109/TIP.2013.2244222.
- MZW15
Kede Ma, Kai Zeng, and Zhou Wang. Perceptual quality assessment for multi-exposure image fusion. IEEE Transactions on Image Processing, 24(11):3345–3356, 2015. doi:10.1109/TIP.2015.2442920.
- NSS15
Mansour Nejati, Shadrokh Samavi, and Shahram Shirani. Multi-focus image fusion using dictionary-based sparse representation. Information Fusion, 25:72–84, 2015. URL: https://www.sciencedirect.com/science/article/pii/S1566253514001213, doi:https://doi.org/10.1016/j.inffus.2014.10.004.
- Pet07
Vladimir Petrovic. Subjective tests for image fusion evaluation and objective metric validation. Information Fusion, 8:208–216, 04 2007. doi:10.1016/j.inffus.2005.05.001.