Summary of the technology
Super-imaging system based on light field camera - Compressed Imaging - MultiView
Project ID : 7-2013-543
During the second half of the 19th century the idea of light field photography was raised by numerous of researchers. The concept is to record more information regarding the scene. Unlike a conventional imaging action that translates the 3-D scene to its 2-D projection, in light field photography some information regarding the light field is recorded as well.
The optical creation of the light field function requires detecting the information distribution among the different field rays per each image pixel. This is commonly done by inserting a lenslet array in front of the sensing array and based on ray model optics to get the various fields in different pixels. In this setup there is a significant investment in pixel count and for achieving M by M light fields per mother pixel, we should increase the sensor pixel count by a factor M along each direction. This is not an acceptable demand.
In this project we develop an efficient light field camera based on compressive imaging capability. A combination of light field photography and compressive imaging is the basis of our research to reduce significantly the required pixel count. This provides the capability to capture the three dimensional image based on much more compact and cheaper camera.
Stage of Development
In the recent years we investigated various configurations for achieving light field photography. We developed a preliminary version of a simulator that simulates the various light field photography setups. Based on the simulator, we obtained some indications regarding the performance of each configuration in terms of light efficiency, spatial resolution and image quality. With respect to the Compressive Imaging, a detailed analysis was done and the appropriate method to be used was selected.
To achieve sufficiently low pixel count, we need to use multi exposure methods. Thus, frame registration is needed. This topic was intensively investigated in our laboratory and we have an efficient way to fuse images adequately. Additional activity was done in gaining knowledge for designing and producing the element. Prior knowledge was achieved in the clean rooms of Tel Aviv University while the team holds the know-how of the production process. Additional aspects of the camera design such as its aperture shaping were investigated as well.
The proposed project provides the capability to capture three dimensional images with a long list of benefits: Changing the focus of the image; Digital re-focusing; All in focus; Changing the point of view; Performing 3-D mapping; Changing the iris; Augmented reality. Unlike conventional light field photography cameras (e.g. Lytro), this is achieved with reasonable camera complexity, price, size, perfect image quality and low light performance. The approach is relevant to the whole uses of cameras and may be commercialized via licensing.