| Summary: | 碩士 === 國立成功大學 === 電腦與通信工程研究所 === 97 === After availability of 3D LCD display systems, people can perceive stereo images without wearing any special 3D glasses. There are many researches focused on 2D to 3D video generations due to lack of 3D video contents. The common 2D to 3D conversion technique first performs the depth map estimation and then transfers the original image associated with the estimated depth information into multiview images. Finally, people can preceive autostereoscopic scenes through 3D displays.
In this thesis, we propose the concept of homogeneity to generate the homogeneous edge map of input images. We then extract the active contour by using Snake algorithm to homogeneous edge map such that we can robustly segment the desired objects from the input images. Finally, we can assign depth by using some prompts depending on the specified objects in the images. In addition, we also implement a depth assignment system which can help the users to assign depth image of single-veiw images effectively. By using the depth map generation system with the proposed techniques, we can achieve 2D to 3D image conversion easily.
|
|---|
