Pixel-Shift Image Capture for Memory-Limited Super-Resolution
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing super-resolution techniques for generating high-resolution images from low-resolution images are limited by memory constraints in digital cameras, leading to insufficient resolution when optical zoom is maximized, especially in multi-frame super-resolution methods.
Innovation Solution
An image capturing apparatus that determines a first region, sets a pixel shift amount, and synthesizes multiple images to enhance perceived resolution beyond individual image limits, using a processor to manage pixel shifting and image capturing times.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If multi-frame super-resolution technique is used to generate high-resolution image, then perceived resolution is improved, but number of image capturing times is limited by memory constraint
Solution Approach 1:
The image capturing process is segmented into two distinct phases: a first capturing mode that captures a first plurality of images with the sensor in initial positions, and a second capturing mode that captures a second plurality of images with the sensor in shifted positions. This segmentation allows the system to process images in manageable batches that fit within memory constraints while still achieving super-resolution through combination of both sets.
Solution Approach 2:
The patent introduces a temporal dimension to the image capturing process by alternating between two capturing modes across multiple cycles. Instead of requiring all images to be captured simultaneously in memory, the system uses time-based sequencing to accumulate sufficient image data while managing memory usage through progressive processing.
2Measurement precision
If optical zoom is maximized to capture desired region, then angle of view is optimized, but resolution is insufficient due to limited image capturing times
Solution Approach 1:
The patent implements dynamic switching between two capturing modes depending on the zoom level. When optical zoom is maximized to capture a specific desired region, the system activates the second capturing mode with pixel shifting to enhance resolution. When zoom is not maximized, the first capturing mode is used. This dynamic adaptation allows the system to optimize both resolution and angle of view based on real-time capturing conditions.
3Measurement precision
If pixel shifting is performed to generate super-resolution image, then perceived resolution is improved, but memory constraint limits the number of pixels that can be used
Solution Approach 1:
The patent performs preliminary processing by generating a super-resolution image from the first plurality of images captured in the first mode before capturing the second plurality of images in the second mode. This preliminary super-resolution image serves as a foundation that reduces the memory burden for subsequent processing, as the system can work with a pre-processed intermediate result rather than managing all raw images simultaneously.
Data Source
AI summary
An image capturing apparatus includes an image sensor configured to capture a plurality of images according to a method for pixel shifting and at least one processor. The at least one processor is configured to determine a first region in each of the plurality of images, determine a number of times of image capturing based on the first region, set a pixel shift amount in the pixel shifting based on the number of times of image capturing, and synthesize the first regions of the plurality of images.