Simultaneous independent information display at multiple depths using multiple projectors and patterns created by epipolar constraint and homography transformation

Generally, the projected image of a video projector is spatially invariant and cannot project different images at different depths. If independent images can be projected at different depths, there is a great potential for new AR or MR information presentation devices. In the previous research, solution using multiple projectors is presented, however, there are problems in insufficient contrast. In the paper, we propose a new algorithm to achieve high contrast projection, by restricting a projection pattern to only symbolic information such as letters, which allows a large freedom on pattern optimization to present the same information. For efficency, we also introduce epipolar constraint-based pattern optimization algorithm, which divides the original 2D problem into 1D problems. In our demonstration, it will be shown that several words are presented at different depths with static video projectors with high contrast.