Image quality evaluation of laser-projection light-field 3D display
Date Issued
2016
Date
2016
Author(s)
Hsu, Chun-Chia
Abstract
Recently, auto-stereoscopic displays are the popular topics. There are many fields which use the auto-stereoscopic display technique. Light field displays are the most important topic in the auto-stereoscopic displays, and the systems require several design modifications to achieve the multi view characteristic of stereoscopic image. In the Master Thesis, we design a laser projection light field 3D display combined with the MEMS laser projectors and light field 3D display. The low cost and scalable system include the advantages of the laser projector and wide view angle. We use the two different diffuser screen and corresponding parameters to build two different systems. We also use the software to simulate two systems and compare the image results with the image results of our real system, and observe how the diffusion angle of the screen affect the image results. The crosstalk value of our system are measured and we compare the results with the other auto-stereoscopic display, and our systems have the better image quality of the crosstalk. We use the subjectivity experiment to define the perceptual depth of human, and we compared the two results with the real condition. The image people saw in the system with smaller are closer to the real object, and we design the experiment to find out how two system offer motion parallax by the number of sub-image. And we estimate how the image close to the full motion parallax when we promote the diffusion angle of screen. In the future research, we can displace the planar diffuser with the curved diffuser and there would be some advantages for curved screen. Or the laser speckle is another research in the laser display field. Fatigue is one of the research topics in the stereoscopic display.
Subjects
light-field 3D display:crosstalk
Type
thesis
File(s)![Thumbnail Image]()
Loading...
Name
ntu-105-R02941092-1.pdf
Size
23.32 KB
Format
Adobe PDF
Checksum
(MD5):380b09f424d39ce69d1e2da2d3d08128