Depth Estimation Data Generation for Pseudo 3D Image Separation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing pseudo 3D image generating devices fail to effectively enhance the sense of separation between background and objects in pseudo stereovision, resulting in a low 3D effect due to inadequate reflection of scene analysis in depth estimation data.
Innovation Solution
A depth estimation data generating apparatus that calculates a composition ratio for basic depth models using pixel statistics, generates an object signal emphasizing concavity and convexity, and compensates it with an offset value to improve the 3D effect by enhancing the separation between background and objects.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If basic depth model images are selected by analyzing scene structure and simply added with object information, then the processing is simple, but the sense of separation between background and object is low resulting in low 3D effect
Solution Approach 1:
The patent applies parameter changes by dynamically adjusting the composition ratio of basic depth models based on scene structure analysis parameters (upper/lower screen high frequency component evaluation values). It also changes the object information parameters by selecting from multiple types (R signal, B signal, or both) and applying offset values to emphasize depth, thereby improving depth estimation accuracy while maintaining processing efficiency through automated parameter optimization
Solution Approach 2:
The patent uses composite materials principle by composing multiple basic depth model images (不同类型的深度模型图像) according to calculated composition ratios to create a composite depth model. This composite approach combines the strengths of different depth models (plane model, convex surface model, concave surface model) to achieve more accurate depth estimation that reflects the actual scene structure, thereby improving the sense of separation between background and objects
2Measurement precision
If multiple types of basic depth models are composed according to composition ratio, then the scene structure estimation is improved, but the object information does not reflect scene analysis
Solution Approach 1:
The patent resolves this contradiction by changing the object information parameters dynamically based on scene analysis results. It selects different object information types (R signal, B signal, or both) according to the composed depth model characteristics and applies offset values calculated from scene structure analysis. This ensures object information accurately reflects the analyzed scene structure while maintaining high measurement precision for depth estimation
Solution Approach 2:
The patent implements feedback by using the composition ratio calculation results (based on upper/lower screen high frequency component analysis) to guide the selection and processing of object information. The scene structure analysis feedback loop ensures that object information is appropriately adjusted to match the composed depth model, preventing information loss and maintaining accuracy in depth estimation
Data Source
AI summary
An RB rate calculator calculates an RB rate based on an R signal and a B signal. A starting point changing unit changes a starting point based on the RB rate. An offset calculating unit calculates an offset value to adjust for selection of a basic depth model type based on a bottom high frequency component evaluation value. An adding unit adds a signal from the starting point changing unit and an offset. Another adding unit adds an offset-added signal from the adding unit and a basic depth model-composed image signal supplied from a composing unit, and generates depth estimation data wherein a degree of superimposition of object information is changed according to a composition of a composed image of basic depth models selected to be composed.