The Computational Photography is a new and rapidly developing subject. By integrating a variety of technologies such as digital sensors, optical systems, intelligent lighting, signal processing, computer vision, and machine learning, computational photography aims at improving the traditional imaging technology, in which an image is formed directly at sensors. The joint force Computational Photography enhances and extends the data acquisition capabilities of traditional digital cameras, and captures the full range of real-world scene information.
With rapidly advancing hardware, some studies use highly curved image sensors to improve optical performance, some try to optimize the micro-lens array-based light field camera systems, the others propose fundamentally new imaging modalities for depth cameras. Despite that new sensing technologies are able to provide better image quality even richer information, cost constraints often limit large scale applications of such technologies. Instead, learning-based computational photography techniques demonstrate a potential capability of enhancing the camera systems without requiring a significate upgrade of hardware. Recently, deep neural networks have shown their superior performance in the imaging computation. They can either learn a complex imaging mechanism in a low-light environment, detect objects to strengthen the focus function and depth estimation, or enhance the degraded images captured in bad work conditions, just to name a few.
The objective of this special issue is to provide a forum for researchers to share their recent progresses on deep learning for computation photography. Papers could cover broad aspects from both theoretical and engineering perspectives, including DNN techniques for modeling, DNN algorithms for image reconstruction, and novel DNN designs for computational imaging in various spectral regimes, such as optical, multi-spectrum, ultrasound, microwave regimes, and so on. Contributions are also welcome concerning applications using computational photography, from fundamental science to applied research.
Potential topics include, but are not limited to:
- Computational imaging methods and models
- Computational illumination
- Computational image processing
- Multi-spectral imaging, SAR imaging, medical imaging and their processing
- Post-processing in computational photography
- Degraded image enhancement
- Aesthetics captioning
- Image recovery from compressed sensing
- Image generation through domain learning
- Applications, including natural, medical, remote sensing research.