3D Image Processor Aspect Conversion Memory Optimization

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing three-dimensional image processing methods require significant memory capacity and suffer from reduced scaling precision when performing 2D/3D conversion and aspect conversion, either due to increased frame memory requirements or decreased signal resolution.

Innovation Solution

A three-dimensional image processor and method that includes a discriminator to determine orientation, a parameter generator for aspect ratio and parallax, and a controller to manage read positions in frame memory, allowing for efficient 2D/3D and aspect conversion without increasing memory capacity and maintaining precision.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Adaptability or versatility

If 2D/3D conversion is performed first and then aspect conversion is performed, then three-dimensional display capability is achieved, but the required capacity of the frame memory doubles

Engineering Contradiction:
Improve2D/3D conversion capabilityVSAvoidframe memory capacity
Core Design Contradiction:
Adaptability or versatilityVSQuantity of substance

Solution Approach 1:

The patent performs aspect conversion before 2D/3D conversion, reversing the conventional order. By pre-converting the aspect ratio of the basic image to match the display device orientation, the subsequent 2D/3D conversion operates on already-resized data, eliminating the need to store both original and converted frames simultaneously in memory.

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The patent inverts the conventional processing sequence by performing aspect conversion first and then 2D/3D conversion. This reversal of the operation order fundamentally changes the memory requirements, as the intermediate results need not be stored in full resolution for both operations.

Inventive Principle:
Principle #13The other way round (Inversion)

2Quantity of substance

If the right-eye signal and the left-eye signal are output collectively in one frame, then the required capacity of the frame memory does not increase, but the resolution of the signals falls

Engineering Contradiction:
Improveframe memory capacityVSAvoidsignal resolution
Core Design Contradiction:
Quantity of substanceVSMeasurement precision

Solution Approach 1:

The patent performs aspect conversion on the basic image before 2D/3D conversion, preparing the image at the correct aspect ratio in advance. This preliminary resizing ensures that when signals are packed into a single frame, the resolution is maintained because the conversion operations work on properly scaled data from the beginning.

Inventive Principle:
Principle #10Preliminary action

3Productivity

If aspect conversion is performed first and then 2D/3D conversion is performed, then the processing order is optimized, but the shift direction of the right-eye signal and the left-eye signal varies depending on orientation

Engineering Contradiction:
Improveprocessing efficiencyVSAvoidorientation adaptability
Core Design Contradiction:
ProductivityVSAdaptability or versatility

Solution Approach 1:

The patent makes the processing parameters dynamic by adjusting the 2D/3D conversion shift direction based on the display device's orientation. The system detects whether the device is in portrait or landscape mode and automatically adjusts the parallax shift direction accordingly, allowing aspect conversion followed by 2D/3D conversion to work correctly in all orientations.

Inventive Principle:
Principle #15Dynamics

Solution Approach 2:

The patent changes the shift direction parameter of the 2D/3D conversion based on the display orientation. When the display device is in portrait mode, the shift direction is adjusted differently than in landscape mode, ensuring correct stereoscopic effect regardless of orientation while maintaining the optimized processing order.

Inventive Principle:
Principle #35Parameter changes

Data Source

PatentUS9350971B2Three-dimensional image processor and three-dimensional image processing method
Publication Date: 2016.05.24 JVC KENWOOD CORP
  • US9350971B2 patent drawing
  • US9350971B2 patent drawing
  • US9350971B2 patent drawing

AI summary

A three-dimensional image processor includes an aspect conversion discriminator, an aspect conversion parameter generator, a 2D/3D conversion parameter generator, a frame memory unit and a read address controller. The aspect conversion discriminator generates a signal indicating whether or not a read position of a video signal is to be modified. The aspect conversion parameter generator generates a parameter indicating an aspect ratio of an image based on the video signal. The 2D/3D conversion parameter generator generates shift amounts for generating a left-eye signal and a right-eye signal based on the video signal. The frame memory unit temporarily stores the video signal. The read address controller controls the read position of the video signal based on the parameter and the shift amounts when the signal received from the aspect conversion discriminator indicates that the read position is to be modified.