Depth Map Generation for Human Figures Using Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional methods for generating depth maps from two-dimensional images often result in unnatural 3D pop-out effects, particularly when displayed on larger screens, and can produce artifacts near boundaries, flicker between frames, and fail to create 3D objects without relative motion.
Innovation Solution
A depth information generating device and method that detect human faces and figures, extracting and separating them from the background using distinct depth values, and employing smoothing processing to enhance edge clarity and reduce flicker, allowing for real-time operation with minimal memory capacity.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If conventional depth map generation methods are used, then processing speed is maintained, but the 3D pop-out effect becomes unnatural and artifacts appear at boundaries
Solution Approach 1:
The patent segments the image into foreground objects and background regions, applying different depth map generation strategies to each segment. This allows precise control over depth values at boundaries, preventing artifacts while maintaining natural 3D pop-out effects.
Solution Approach 2:
The patent applies local quality adjustment by treating foreground and background regions differently. Foreground objects receive enhanced depth processing with object-aware algorithms, while background regions use conventional methods, optimizing overall depth map quality without uniform processing overhead.
2Ease of operation
If face-only extraction is used, then processing complexity is reduced, but the 3D pop-out effect is insufficient for the entire human body
Solution Approach 1:
The patent performs preliminary face detection to establish a starting point, then automatically expands the selection to the entire human body using body part detection algorithms. This preliminary action simplifies the interface while achieving complete body 3D effects.
Solution Approach 2:
The patent creates a multi-functional system that can detect and process various human body parts (face, body, limbs) using a unified depth map generation framework. This allows the same processing pipeline to handle both simple face cases and complete body cases.
3Manufacturing precision
If foreground extraction based on color segmentation is used, then depth separation is achieved, but flicker occurs between video frames
Solution Approach 1:
The patent maintains continuity of useful action by tracking detected objects across video frames and preserving their depth assignments. This ensures consistent foreground-background separation without flicker, as the system continuously refines rather than re-detects object boundaries.
Solution Approach 2:
The patent implements feedback mechanisms where depth map results from previous frames inform current frame processing. This feedback loop stabilizes object detection across frames, reducing flicker while maintaining accurate foreground-background separation.
Data Source
AI summary
A depth information generating device includes a region extracting unit that detects a human face in at least one two-dimensional image, and based on the detected face, extracts a human figure region indicating a human figure within a region of the at least one two-dimensional image; and a depth map generating unit that gives a depth value different from a depth value of a region other than the human figure region to the human figure region to generate a depth map that separates the human figure region from the region other than the human figure region.