Image Encoding via Gaze-Based Importance Factor
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional image encoders face limitations in efficiently encoding images due to the complexity and delay caused by identifying regions of interest and using extra bits, which restricts the number of bits that can be employed.
Innovation Solution
A system and method that identify and eliminate non-important motion by calculating an importance factor based on the extent of change and distance from the gaze location, allowing for the reuse of previous encoded data for non-important areas.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional encoders identify regions of interest and employ extra bits to encode such regions, then encoding precision is improved, but device complexity increases and processing time increases
Solution Approach 1:
The patent applies local quality by differentiating encoding strategies based on spatial location relative to the gaze point. Areas closer to the gaze point receive higher encoding precision while areas farther away receive reduced precision, eliminating the need for uniform extra bits across the entire image and thus reducing overall device complexity
Solution Approach 2:
The patent segments the image into multiple regions based on distance from the gaze point and applies different encoding parameters to each segment. This segmentation allows the encoder to allocate bits more efficiently across different regions, improving overall encoding precision without uniformly increasing complexity throughout the entire image
2Measurement precision
If conventional encoders employ extra bits to encode regions of interest, then encoding precision is improved, but processing time increases
Solution Approach 1:
By applying local quality, the patent concentrates encoding resources on regions near the gaze point where precision is most critical, while reducing processing time in distant regions where human perception is less sensitive to changes, thus achieving a better time-precision tradeoff
Solution Approach 2:
The patent applies partial action by encoding only the necessary portions of the image with high precision rather than uniformly processing the entire image. This selective encoding reduces overall processing time while maintaining sufficient precision for the most important regions
3Loss of time
If conventional encoders use limited bits for encoding, then processing time is reduced, but encoding precision deteriorates
Solution Approach 1:
The patent resolves this contradiction by applying local quality - using more bits for regions near the gaze point where precision is most important and using fewer bits for distant regions, thereby achieving high overall encoding precision without requiring excessive total bits that would increase processing time
Solution Approach 2:
The patent changes encoding parameters dynamically based on the distance of each region from the gaze point. Regions closer to the gaze point use higher bit depths and more sophisticated encoding parameters, while distant regions use reduced parameters, optimizing the balance between processing time and encoding precision
Data Source
AI summary
A gaze location is identified in a given image, based on a given gaze direction. The given image is divided into a plurality of areas. For a given area of the given image, a corresponding area is identified in at least one previous image. An extent of change is determined between the corresponding area of the at least one previous image and the given area of the given image. An importance factor is then calculated for the given area of the given image, based on the determined extent of change and a distance of the given area from the gaze location. The given image is encoded into encoded image data. When the importance factor for the given area is smaller than a first predefined threshold, the step of encoding comprises re-using previous encoded data of the corresponding area, instead of encoding replacing the given area in of the given image with the corresponding area into the encoded image data.