Super Resolution Digital Photography Using Neighborhood Patch Stitching
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Modern mobile devices with embedded cameras face issues such as jagged edges in scaled images and limited field-of-view, requiring users to take multiple overlapping shots for panoramic images, which is cumbersome and artifact-prone.
Innovation Solution
The system generates super-resolution digital images and extends the field-of-view by selecting image regions, computing distance measurements, determining closest neighborhood patches, and stitching scaled portions from previously captured images to create seamless, high-quality images beyond the camera's fixed field-of-view.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If digital scaling is used to create cropped images, then a close-up view of an object is obtained, but jagged edges and undesirable artifacts are generated
Solution Approach 1:
The system captures multiple images at different distances/positions in advance (first image at first distance, second image at second distance closer to object). These preliminary images are stored and later used to reconstruct the cropped region, avoiding the need for digital scaling that causes artifacts.
Solution Approach 2:
Instead of digitally scaling an existing image, the system copies and stitches actual captured image data from the first image to create the cropped view. This uses real optical information rather than interpolated pixels, eliminating jagged edges.
2Area of stationary object
If a digital camera with fixed field of view is used, then the camera structure is simple, but the user cannot capture an entire scene in a single shot
Solution Approach 1:
The system automatically captures a first image at a first distance before the user takes the second image at a closer distance. This preliminary capture stores scene information that will be used to extend the field of view, eliminating the need for manual panoramic photography.
Solution Approach 2:
The system automatically performs the field of view extension by retrieving the first image, determining the extended portion, extracting it, scaling it, and stitching it to the second image. This eliminates the need for users to manually take multiple overlapping shots and stitch them themselves.
3Area of stationary object
If the user takes multiple overlapping shots for panoramic images, then the entire scene can be captured, but the process is cumbersome and requires manual stitching
Solution Approach 1:
The system automatically retrieves the first image using timestamp and FOV direction data, determines the extended portion, extracts it, scales it to match the second image, and stitches them together. This automated process eliminates the need for users to manually take multiple shots and perform stitching operations.
Solution Approach 2:
The system merges the first image and second image by extracting the extended portion from the first image and stitching it to the second image. This combines multiple captured images into a single extended field of view image automatically.
Data Source
AI summary
Techniques are disclosed for creating scaled images with super resolution using neighborhood patches of pixels to provide higher resolution than traditional interpolation techniques. Also disclosed are techniques for creating super field-of-view (FOV) images of a scene created from previously captured and stored images of the scene that are stitched together with a current image of the scene, to generate an image of the scene that extends beyond the fixed FOV of the camera.