Omnidirectional Image Encoding Resolution Distribution
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing image processing technologies, such as equidistant cylindrical projection, result in high pixel density near the poles of celestial sphere images, leading to inefficient encoding and decoding, especially since humans primarily focus on horizontal directions, making these areas less important.
Innovation Solution
An image processing device and method that encodes omnidirectional images with a middle image at a higher resolution and top and bottom images at lower resolutions, allowing for efficient encoding and decoding while maintaining image quality by reducing pixel density near the poles.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If equidistant cylindrical projection is used to generate celestial sphere image, then the image can be mapped to sphere surface, but pixel density increases near the poles leading to inefficient encoding
Solution Approach 1:
The patent applies local quality by dividing the celestial sphere image into three vertical regions (top, middle, bottom) and encoding each region at different resolutions. The middle region corresponding to the horizontal direction maintains high resolution, while the top and bottom regions near the poles use lower resolution, optimizing encoding efficiency based on local importance.
Solution Approach 2:
The patent segments the omnidirectional image into multiple divided images vertically (top, middle, bottom sections) before encoding. This segmentation allows different resolution settings for different regions, with the middle section encoded at first resolution and top/bottom sections encoded at second resolution, resolving the contradiction between mapping accuracy and encoding efficiency.
2Manufacturing precision
If high resolution is maintained for entire celestial sphere image, then image quality is preserved, but encoding and decoding time increases
Solution Approach 1:
The patent encodes the middle divided image at high first resolution to preserve image quality in the horizontally-focused region, while encoding top and bottom divided images at lower second resolution. This selective quality approach reduces total encoding/decoding time while maintaining perceptual image quality where it matters most.
Solution Approach 2:
The patent applies partial action by providing high resolution only where needed (middle horizontal region) rather than uniformly across the entire image. The top and bottom regions receive reduced resolution, reducing overall processing time while maintaining sufficient quality for the important central region.
3Device complexity
If uniform resolution is used for all divided images, then encoding is simplified, but pixel density near poles becomes excessively high
Solution Approach 1:
The patent uses local quality by assigning different resolutions to different vertical regions of the image. The middle region uses first resolution while top and bottom regions use second resolution, reducing pixel density near the poles where it is excessively high, while maintaining appropriate density in the horizontal region.
Solution Approach 2:
The patent changes the resolution parameter selectively across different regions of the image. By varying the resolution parameter (first resolution for middle, second resolution for top/bottom), the patent optimizes pixel density distribution and reduces encoding complexity compared to uniform high-resolution encoding.
Data Source
AI summary
An image processing device and an image processing method for generating a celestial sphere image are provided such that the pixels near the poles of the sphere are kept from increasing in density when the image is mapped to the sphere surface. An encoder encodes, with respect to an omnidirectional image generated by equidistant cylindrical projection to include a top image, a middle image, and a bottom image in a vertical direction, the middle image into an encoded stream at a high resolution, and the top image and the bottom image into encoded streams at a resolution lower than the high resolution in systems including image display systems.