3D Image Rendering with Depth Remapping for Clearer View Synthesis

The patent applies preliminary action by pre-calculating and storing depth information for multiple viewpoints in advance. Instead of converting 2D to 3D in real-time during playback, the system pre-processes the video content to generate depth maps and multi-viewpoint data, which are then readily available for high-quality 3D rendering without degradation.

The patent introduces depth information as an intermediary element between the 2D video and the 3D display output. By incorporating depth maps and depth buffers as intermediate data structures, the system enables accurate 3D rendering through view synthesis techniques, avoiding the quality loss associated with direct 2D-to-3D conversion methods.

The patent replaces traditional mechanical 2D-to-3D conversion methods with a computational approach based on depth information and view synthesis. Instead of using optical or mechanical conversion techniques that inherently cause blurring, the system uses algorithms that leverage pre-calculated depth data to synthesize accurate 3D views, maintaining sharp image quality.

The patent changes the fundamental parameters of the video data by incorporating depth information (Z-axis parameter) alongside traditional color information. This transformation from 2D (x, y, color) to 3D (x, y, z, color) parameter space enables precise control over image rendering at different viewpoints without the blurring artifacts that plague conventional conversion methods.