Omnidirectional Image Distribution with Variable Pixel Density
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional image distribution systems that superimpose high-resolution images on low-resolution panoramic images can create a strange viewing experience due to visible boundaries, leading to excessive information transmission.
Innovation Solution
A method that generates a distribution image by thinning out and extracting pixels from omnidirectional images, using a function that increases pixel intervals from the center to the edge, arranging pixels within a circular region based on the field of view and user's line of sight, to create a seamless transition in image quality when expanded to a panoramic view.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If high-resolution images are superimposed on low-resolution panoramic images, then image quality at the center is improved, but visible boundaries appear and user strangeness increases
Solution Approach 1:
The patent applies local quality by differentiating pixel extraction density across different regions of the distribution image. High-density pixel extraction is performed in the central region where users focus attention, while low-density extraction is applied to peripheral regions. This creates a gradient in image quality that matches human visual attention patterns, eliminating abrupt boundaries while maintaining center quality.
Solution Approach 2:
Instead of uniformly distributing high-resolution pixels throughout the entire image, the patent inverts the conventional approach by concentrating pixel density where it matters most (the center) and sparsifying it toward the edges. This inversion of the typical quality distribution strategy eliminates the unnatural boundary effect while preserving essential image information.
2Measurement precision
If high-resolution images are superimposed on low-resolution panoramic images, then image quality is improved, but the amount of information for video distribution increases
Solution Approach 1:
The patent segments the distribution image into different quality zones based on spatial location. The central region receives high-resolution pixel data while peripheral regions receive low-resolution data. This segmentation allows the system to transmit only the necessary amount of information to each region, optimizing the balance between image quality and data volume.
Solution Approach 2:
By assigning different quality levels to different spatial regions, the patent reduces the overall information quantity required for transmission. Instead of transmitting high-resolution data for the entire image, only the central high-quality region is transmitted at full resolution, significantly reducing the total data size while maintaining perceived image quality.
3Ease of manufacture
If uniform pixel extraction is applied across the entire distribution image, then processing is simplified, but image quality decreases at the center and user experience deteriorates
Solution Approach 1:
The patent implements local quality through a systematic pixel extraction function that varies extraction density by spatial position. The function calculates the distance of each pixel from the center and applies different extraction criteria accordingly. This adds only minimal computational complexity while dramatically improving central image quality and user experience.
Data Source
AI summary
A method for generating a distribution image to be distributed via a network includes acquiring an image taken at least omnidirectionally, and generating a distribution image having a low image-quality part at four corners by thinning out and extracting pixels from the acquired image. The generating includes extracting the pixels by selecting the pixels to be extracted in accordance with a ratio between a circle having a center of the distribution image as an origin and an arc of the circle included in the distribution image.