Image Sensor Missing Pixel Compensation via Focused and Defocused Capture
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Image sensors with missing pixels, either defective or performing non-image sensing functions, degrade image quality and can lead to improper object recognition, especially when these pixels are numerous.
Innovation Solution
A method and device that generate a final image by compensating data for missing pixels using both focused and defocused images, where data from neighboring pixels is used to fill in missing pixel data, improving image quality and reliability.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If pixels for non-visible light sensing functions (e.g., infrared filter) are included in the image sensor, then additional sensing capabilities are provided, but the quality of the generated visible light image is degraded
Solution Approach 1:
The patent separates the image capture process into two distinct phases: focused capture for visible light image quality and defocused capture for obtaining data from non-visible light sensing pixels. This segmentation allows each phase to serve its specific purpose without compromising the other, resolving the contradiction between versatility and image quality.
Solution Approach 2:
The patent performs preliminary defocused capture to obtain data from non-visible light sensing pixels before final image processing. This preliminary action allows the system to gather additional sensing information without interfering with the primary visible light image quality, as the defocused data is processed separately and integrated only where needed.
2Adaptability or versatility
If one or more pixels for non-image sensing functions are included in the image sensor, then multi-functionality is achieved, but the generated image may not properly recognize the object
Solution Approach 1:
The patent dynamically adjusts the focus state of the lens system between focused and defocused positions. By switching between these states, the system can selectively capture visible light images with proper object recognition while also gathering data from non-visible light sensing pixels, thus maintaining both reliability and versatility.
Solution Approach 2:
The patent uses defocused image data as an intermediary to bridge the gap between non-visible light sensing pixels and the final visible light image. The defocused capture allows data from non-visible light pixels to be extracted and integrated without directly compromising the visible light image quality, serving as a mediator that preserves object recognition accuracy.
3Manufacturing precision
If data from neighboring pixels is used to compensate missing pixel data, then image quality is improved, but processing complexity increases
Solution Approach 1:
The patent changes the focus parameter of the lens system between focused and defocused states to capture images under different conditions. This parameter change enables the system to obtain data from non-visible light sensing pixels through defocused capture, and by comparing focused and defocused images, compensate for missing pixel data while managing processing complexity through structured comparison algorithms.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
The approach results in a higher quality and reliable final image by effectively incorporating data from neighboring pixels, even when mathematical inference is difficult, and reduces the impact of missing pixels on image recognition.
Implementation Method 1
a lens configured to receive light reflected from an object
Implementation Method 2
generating an imperfect image excluding data corresponding to one or more pixel sensors by capturing an object while the object is focused
Implementation Method 3
The photo-sensitive device corresponding to the filter generates an electrical signal based on intensity of the visible light which passed the filter
Data Source
AI summary
A method of generating an image using an image capturing device includes generating an imperfect image excluding data corresponding to one or more pixel sensors by capturing an object while the object is focused. A defocused image is generated to include the data corresponding to the one or more pixel sensors generated using neighboring pixel sensors around the one or more pixel sensors while the object is defocused. The data corresponding to the one or more pixel sensors is extracted based on data in the generated imperfect image and data in the generated defocused image. A final image is then generated by reflecting the extracted data to at least one of the generated imperfect image or the generated defocused image.