2D to Multi-Viewpoint Image Conversion for Naked-Eye 3D Displays
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for converting a two-dimensional image into a multi-viewpoint image in naked-eye 3D displays are complex and prone to defects such as image cracks and distortions, making them difficult to implement effectively.
Innovation Solution
A display method and system that preprocesses two-dimensional images by calculating depth values and generating layered images, interpolating blank areas, and assembling single-viewpoint images to form a synthesized multi-viewpoint image, which is then processed for output to a naked-eye 3D display screen, using techniques like sharpening, Fourier transformation, and interpolation algorithms to reduce distortions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If conventional methods are used to convert single-viewpoint content into multi-viewpoint images, then multi-viewpoint display can be achieved, but the process becomes complicated and produces image defects such as cracks, distortion and jitter
Solution Approach 1:
The patent segments the image processing into distinct modules: a layer separation module that divides the image into multiple depth layers, a multi-viewpoint generation module that creates different viewpoint images from each layer, and a synthesis module that combines them. This segmentation simplifies the overall conversion process while maintaining image quality by handling each aspect separately with specialized algorithms.
Solution Approach 2:
The patent performs preliminary depth estimation and layer separation on the single-viewpoint image before generating multi-viewpoint images. By pre-calculating depth information and organizing pixels into depth-based layers beforehand, the system avoids complex real-time computations during viewpoint generation, reducing both computational complexity and image artifacts.
2Adaptability or versatility
If depth information is estimated from a single-viewpoint image, then multi-viewpoint display is enabled, but image cracks and distortions occur due to inaccurate depth reconstruction
Solution Approach 1:
The patent introduces depth layers as an intermediary structure between the single-viewpoint input and multi-viewpoint output. By estimating depth information and organizing pixels into intermediate depth layers, the system creates a structured representation that facilitates accurate multi-viewpoint generation while reducing artifacts like cracks and distortions that occur with direct conversion methods.
3Reliability
If multiple images with multiple viewpoints are displayed for naked-eye 3D TV, then genuine 3D effect is achieved, but the content source is limited and conversion from single-viewpoint is required
Solution Approach 1:
The patent adds a depth dimension to traditional 2D single-viewpoint images by estimating depth information and creating depth-based layers. This dimensional transformation enables the system to generate multiple viewpoint images from single-viewpoint content, expanding content compatibility while maintaining the naked-eye 3D display effect that requires multi-viewpoint input.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
The method simplifies the conversion process, reduces image distortions, and provides a stable multi-viewpoint image display for naked-eye 3D technology, making it easier to use and cost-effective.
Implementation Method 1
calculating a depth value of the target object according to a frequency component of the target object
Data Source
AI summary
Display method and system for converting 2D image into multi-viewpoint image is disclosed, comprising: acquiring and tagging a target object within a 2D image; calculating a depth value according to a frequency component; generating a layered image before viewing from different preset viewpoints; tagging a viewpoint image; estimating before filling a pixel in a blank area of a virtual viewpoint image, based on a depth value difference of the layered image, generating and saving sequentially a single-viewpoint image output, before detecting and filling a blank area in it; detecting before smoothing a sudden change area; assembling to form a synthesized image, processing and sending to a naked-eye 3D display screen for displaying. It converts a 2D image to a multi-viewpoint image, provides a naked-eye 3D display, reduces image distortion, easy and convenient to use, with a low cost.