ML Upscaling Frame Edge Padding for Sharp Boxing Borders
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional ML-based upscaling processes produce artifacts along image edges when boxing edges are present, leading to loss of clear and sharp edges in the final output.
Innovation Solution
Process frame edges in the input image prior to ML-based training by detecting and extending boxing edge regions with extra pixels, upscaling these regions, cropping them to the original size, and inserting them into the output image, while handling different types of edge and corner sub-regions using specific padding techniques.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If conventional ML-based upscaling is applied to images with boxing edges, then the upscaling process can be performed, but artifacts are produced along the edges leading to loss of clear and sharp edges
Solution Approach 1:
The patent applies preliminary action by detecting boxing edges and extending the image with padding pixels before the ML upscaling process. This preprocessing step prepares the image by adding extra pixels around boxing edge regions, which prevents artifacts from forming during subsequent upscaling operations.
Solution Approach 2:
The patent segments the image processing into distinct regions: boxing edge regions are identified and separated from the main image content. Different processing strategies are applied to edge regions versus interior regions, with edge regions receiving specialized padding and extension treatment to preserve sharpness and prevent artifacts.
2Manufacturing precision
If boxing edge regions are extended with extra pixels, then edge artifacts are reduced, but the processing complexity increases due to multiple steps (detection, extension, upscaling, cropping, insertion)
Solution Approach 1:
The patent merges multiple processing operations into an integrated pipeline: edge detection, padding extension, ML upscaling, and result integration are combined into a unified processing flow. This allows the system to handle edge regions and interior regions in a coordinated manner, reducing overall complexity compared to separate independent processing steps.
Data Source
Figure 1A~1C
Figure 2
Figure 3
AI summary
Frame handling an input image in an ML-based upscaling process to produce an output image, including: detecting and extracting boxing edge regions from edges of an active picture area of the input image; extending the extracted boxing edge regions by adding extra pixels around each of the boxing edge regions to produce padded edge regions; upscaling the padded edge regions to produce upscaled padded edge regions; cropping each of the upscaled padded edge regions to a size corresponding to the size of each of the boxing edge regions prior to the upscaling multiplied by an upscaling factor; inserting the cropped edge regions into the edges of the output image; and upscaling a non-edge region of the input image and inserting the upscaled non-edge region into the output image.