Imaging Pixel Compensation for High Spatial Frequency Artifacts
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing imaging apparatuses that use a pupil-division type phase difference detection method for focus detection suffer from decreased image quality when the spatial frequency approaches the Nyquist frequency, particularly when focus detection pixels are arranged linearly, leading to noticeable image deterioration in high spatial frequency components.
Innovation Solution
The imaging apparatus includes two-dimensionally arranged imaging pixels and non-imaging pixels, with a continuity detection portion that determines the direction of pixel signal continuity and performs specific calculations for non-imaging pixels based on the detected direction, using averaging, interpolation, and spectral distribution to compensate pixel signals, thereby maintaining image quality even at high spatial frequencies.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If focus detection pixels are arranged amongst the imaging pixels to enable pupil-division type phase difference detection, then focus detection capability is improved, but image quality deteriorates at high spatial frequencies due to increased pixel pitch and reduced Nyquist frequency
Solution Approach 1:
The patent applies local quality by differentiating the compensation method based on the directional characteristics of pixel arrangements. When non-imaging pixels are arranged linearly, the system detects the continuity direction and applies directional compensation only in the perpendicular direction, rather than applying uniform compensation in all directions. This localized approach preserves high spatial frequency information in the continuity direction while correcting artifacts in the perpendicular direction.
Solution Approach 2:
The patent changes the compensation parameters dynamically based on the detected continuity direction. The system calculates different compensation weights and interpolation factors depending on whether the non-imaging pixels are arranged horizontally, vertically, or diagonally. This parameter adaptation allows the system to optimize compensation for each specific arrangement configuration, resolving the contradiction between focus detection and high-frequency image quality.
2Reliability
If averaging process is used to compensate pixel data at focus detection pixel positions, then focus detection is enabled, but white and black are inversely imaged at Nyquist frequency due to pitch increase
Solution Approach 1:
The patent applies inversion by detecting the continuity direction and applying compensation in the opposite (perpendicular) direction. Instead of trying to restore the original pixel pitch in all directions, the system identifies where the pitch distortion occurs (perpendicular to the non-imaging pixel arrangement) and applies targeted compensation only there, while leaving the continuity direction unchanged. This inverted approach correctly handles the inverse imaging problem at Nyquist frequency.
Solution Approach 2:
The patent makes the compensation process dynamic by first detecting the continuity direction through correlation analysis of pixel signals, then adapting the compensation method accordingly. The system dynamically selects between different compensation strategies based on the detected arrangement pattern, allowing flexible handling of various spatial frequency conditions and preventing information loss in high-frequency regions.
3Device complexity
If linear arrangement of focus detection pixels is used, then phase difference detection is simplified, but image deterioration becomes visibly noticeable in high spatial frequency components
Solution Approach 1:
The patent applies segmentation by dividing the compensation process into independent directional components. The system separates the compensation operation into the continuity direction (where no compensation is needed) and the perpendicular direction (where compensation is applied). This segmentation allows the linear arrangement simplicity to be preserved while adding targeted compensation only where necessary, maintaining focus detection simplicity while improving high-frequency image quality.
Data Source
AI summary
An imaging apparatus includes an imaging element having two-dimensionally arranged imaging pixels that receive light of an image transmitted through an optical system and output pixel signals corresponding to the received light, and in which non-imaging pixels that are different from the imaging pixels are arranged. The apparatus detects a direction of continuity of the pixel signals based on pixel signals of the imaging pixels positioned around the non-imaging pixels, and determines, based on the detected direction of continuity, the pixel signals at the positions of the non-imaging pixels by processing the pixel output of the imaging pixels positioned around the non-imaging pixels.