We create a system for semi-automatically converting unconstrained 2D images and videos into stereoscopic 3D. Current efforts are done automatically or manually by rotoscopers. The former prohibits user intervention, or error correction, while the latter is time consuming, requiring a large staff. Semi-automatic mixes the two, allowing for faster and accurate conversion, while decreasing time to release 3D content. User-defined strokes for the image, or over several keyframes, corresponding to a rough estimate of the scene depths are defined. After, the rest of the depths are found, creating depth maps to generate stereoscopic 3D content, and Depth Image Based Rendering is employed to generate the artificial views. Here, depth map estimation can be considered as a multi-label segmentation problem, where each class is a depth value. Optionally, for video, only the first frame can be labelled, and the strokes are propagated using a modified robust tracking algorithm. Our work combines the merits of two respected segmentation algorithms: Graph Cuts and Random Walks. The diffusion of depths from Random Walks, combined with the edge preserving properties from Graph Cuts is employed to create the best results possible. Results demonstrate good quality stereoscopic images and videos with minimal effort.
In this paper, we present a logo and trademark retrieval system for unconstrained color image databases that
extends the Color Edge Co-occurrence Histogram (CECH) object detection scheme. We introduce more accurate
information to the CECH, by virtue of incorporating color edge detection using vector order statistics. This
produces a more accurate representation of edges in color images, in comparison to the simple color pixel
difference classification of edges as seen in the CECH. Our proposed method is thus reliant on edge gradient
information, and as such, we call this the Color Edge Gradient Co-occurrence Histogram (CEGCH). We use this
as the main mechanism for our unconstrained color logo and trademark retrieval scheme. Results illustrate that
the proposed retrieval system retrieves logos and trademarks with good accuracy, and outperforms the CECH
object detection scheme with higher precision and recall.