Light Field Image Downsampling via Pixel Group Aggregation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional downsampling techniques are not well-suited for light field images, as they fail to effectively combine disparately located pixels in the sensor array, resulting in reduced refocusing capability and manipulation of light field images.
Innovation Solution
The proposed solution involves downsampling groups of pixels, such as disks of pixels, rather than individual pixels, by aggregating values for pixels with similar relative positions within adjacent disks or regions, while maintaining spatial adjacencies to reduce spatial resolution without impacting angular resolution, thus preserving the ability to refocus and manipulate the image.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If conventional downsampling techniques are applied to light field images, then the image resolution is reduced, but the refocusing capability and manipulation ability are lost
Solution Approach 1:
The patent segments the light field image data into discrete angular samples (rays) that are systematically organized and downsampled. Instead of treating the entire image as a continuous block, the method divides the angular dimension into discrete samples that can be independently processed while maintaining their spatial relationships, thereby preserving refocusing capability during downsampling.
Solution Approach 2:
The patent transforms the light field image from a conventional 2D spatial representation to a 4D representation (2D spatial coordinates plus 2D angular coordinates). This dimensional expansion allows the downsampling operation to selectively reduce resolution in spatial dimensions while maintaining angular resolution, thus preserving refocusing capability that would be lost in conventional 2D downsampling.
2Reliability
If high-resolution light field images are captured, then the refocusing capability is preserved, but the storage and transmission requirements increase significantly
Solution Approach 1:
The patent extracts only the essential angular information needed for refocusing by systematically downsampling the angular samples. Instead of storing or transmitting the complete high-resolution light field data, the method extracts a reduced set of angular samples that retain the refocusing capability while significantly reducing the overall data volume for storage and transmission.
Solution Approach 2:
The patent changes the sampling parameters of the light field image by adjusting the density of angular samples. By modifying the angular sampling rate and selectively retaining only the necessary samples, the method reduces data volume while maintaining the fundamental refocusing capability that depends on angular information.
3Device complexity
If pixels are combined in conventional downsampling, then the computational complexity is reduced, but the disparately located pixels in light field images cannot be effectively combined
Solution Approach 1:
The patent introduces an intermediary coordinate system that maps the disparately located pixels to a unified angular-spatial representation. This intermediary framework allows pixels from different spatial locations but same angular direction to be systematically identified and combined, solving the problem of effectively combining non-adjacent pixels in light field images.
Solution Approach 2:
By adding angular dimensions to the conventional spatial coordinate system, the patent creates a 4D coordinate framework that systematically organizes pixels. This dimensional enhancement allows the processing algorithm to efficiently identify and combine pixels that are disparately located in spatial dimensions but correspond to the same angular direction, thereby reducing processing complexity while maintaining effectiveness.
Data Source
AI summary
According to various embodiments of the invention, improved downsampling techniques are employed, which can be applied to light field images and which preserve the ability to refocus (and otherwise manipulate) such images. Groups of pixels, rather than individual pixels, are downsampled; such groups of pixels can be defined, for example, as disks of pixels. Such downsampling is accomplished, for example, by aggregating values for pixels having similar relative positions within adjacent disks (or other defined regions or pixel groups) of the image. When applied to light field images, the downsampling techniques of the present invention reduce spatial resolution without sacrificing angular resolution. This ensures that the refocusing capability of the resulting light field image is not reduced and/or adversely impacted.