Pseudo 3D Image Generation via Composite Depth Model Blending

The patent segments the depth information generation process into multiple basic depth models (first, second, and third basic depth models with different depth characteristics). Each model can be independently adjusted through control signals, allowing scene-specific optimization while maintaining efficient automatic processing through the combination unit that blends these segmented models based on calculated composite percentages.

The patent introduces dynamic adjustability by enabling control signals to modify the composite percentages of basic depth models in real-time based on scene characteristics. The combination unit dynamically calculates and applies different weighting factors for each basic depth model, transforming a static automatic processing system into a dynamic one that adapts to varying scene requirements.

The combination unit serves multiple functions: it automatically calculates composite percentages, processes control signals for scene-specific adjustments, blends multiple basic depth models, and generates final depth information. This multi-functional design improves depth accuracy through scene-specific control while avoiding the need for separate dedicated components for each function, thereby limiting the increase in device complexity.

The control signal acts as an intermediary that carries scene-specific adjustment information from external sources to the combination unit. This intermediary mechanism enables precise control over depth model composition without requiring direct modification of the core processing architecture, thus improving accuracy while maintaining relatively simple system structure.