Abstract

We used DIBR (Depth Image Based Rendering) conversion technology to generate multiple virtual images by a specific sub-pixel corresponding to architecture of multi-view image. The multi-view generation system can be divided into four sessions. We use reference image and corresponding depth image to generate the multiple virtual-views through DIBR conversion technology. When the input depth image is not precise enough, causing the edges of objects misplaced, by adding the object edge restoration algorithm to raise the efficiency of hole-filling. Then, we will enhance the multi-view image quality by effectively fill the hole in the image. Finally, we can synthesize the multi-view virtual image to show a three-dimensional verisimilitude effect multi- view images. Our method adopts edge restoration algorithm to enhance the quality of multi-view images. Besides, we design a hardware architecture to implement multi-view generation system for real time processing, and thus can be applied to portable electronic devices.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call