Refocusable Image Capture via Grid Segmentation and Sharpness Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing image capture devices lack the ability to refocus images without specialized hardware, making refocus effects expensive and incompatible with standard cameras, and existing solutions often require dedicated light field cameras or hardware add-ons.
Innovation Solution
A method and apparatus that segment a field of view into grids, determine sharpness levels, identify regions, and capture images for each region to create a refocusable image set, allowing users to selectively focus after capturing, using techniques like panoramic, HDR, or 3-D modes, and storing in specific file formats with metadata.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If dedicated light field cameras or hardware add-ons are used to achieve refocus effect, then the refocus capability is improved, but the device cost and complexity increase significantly
Solution Approach 1:
The patent creates a software-based copy of the refocus effect that normally requires specialized hardware. By using multiple images captured at different focal distances and processing them through depth map generation and blending algorithms, the system replicates the light field refocus capability on standard cameras without needing dedicated plenoptic cameras or hardware add-ons.
Solution Approach 2:
The patent replaces the mechanical/optical hardware system (micro lens arrays, specialized sensors) with a computational software system. The refocus effect is achieved through image processing operations including depth map generation, region identification, and selective blending rather than through specialized optical components.
2Reliability
If multiple images at different focal distances are captured to create refocus effect, then the refocus quality is improved, but the number of shots and time required increase
Solution Approach 1:
The system performs preliminary actions by capturing a set of images at different focal distances in advance. These pre-captured images are then processed to generate depth maps and identify focus regions, enabling refocus operations to be performed quickly on already-captured data rather than requiring additional capture time during the refocusing phase.
Solution Approach 2:
The patent captures more images than strictly necessary (excessive action) to ensure sufficient depth information is available for high-quality refocusing. By capturing a comprehensive set of focal distance variations, the system ensures that adequate source material exists for generating accurate depth maps and achieving superior refocus quality across different focus points.
3Measurement precision
If hardware solutions or sensor-level information are used, then the refocus effect accuracy is improved, but the device compatibility is reduced
Solution Approach 1:
The patent creates a universal software solution that can be deployed across multiple device types including smartphones, tablets, and computers. The system processes standard image formats and works with conventional camera hardware, making the refocus capability adaptable to various devices without requiring specialized sensors or hardware modifications.
Solution Approach 2:
The patent introduces an intermediary software layer that sits between the standard camera hardware and the refocus processing. This intermediary captures standard images, processes them through depth estimation algorithms, and generates the refocused output, thereby bridging the gap between conventional cameras and advanced refocus functionality without requiring direct hardware integration.
Data Source
AI summary
The various embodiments of the present invention disclose an image processing method and system for capturing a refocusable image set. The method comprises of segmenting a field of view into a plurality of grids, determining a sharpness level associated with each of the plurality of grids, identifying one or more regions based on the determined sharpness level of each of the plurality of grids, each region comprising one or more of the plurality of grids and capturing an image associated with each of the identified one or more regions to form the refocusable image set. The method further comprises of detecting a refocusing action on an object, in the field of view, having an associated illumination level and varying the associated illumination level of the object based on the detection.