Image Processing Device Defect Correction Circuit
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing imaging devices face challenges in correcting defects and reducing noise in digital images, particularly when defects change with conditions like temperature and analog gain, leading to deteriorated image quality due to incorrect defect determination and noise reduction effects.
Innovation Solution
An image processing device with a defect correcting unit and a noise-reduction processing unit that share a line memory, using a contrast determining unit and averaging unit to select between correction values based on signal value comparisons and thresholds, switching between averaging and replacement data processing to minimize the influence of defects on noise reduction.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If a common line memory is shared between defect correcting unit and noise-reduction processing unit, then circuit size is reduced, but image quality deteriorates due to wrong defect determination affecting noise reduction processing
Solution Approach 1:
The patent segments the processing flow into distinct stages: defect determination stage and noise reduction stage. The defect determination unit operates first to identify defective pixels, then the noise reduction unit processes non-defective pixels. This segmentation prevents defective pixel data from contaminating the noise reduction processing, thereby maintaining image quality while still sharing the common line memory resource.
Solution Approach 2:
The patent applies preliminary defect determination before noise reduction processing. By identifying and flagging defective pixels in advance, the system prepares the data in a suitable state for subsequent noise reduction processing. This preliminary action ensures that only valid pixel data undergoes noise reduction, preventing quality deterioration while utilizing the shared memory efficiently.
2Productivity
If defect correction and noise reduction are performed in parallel, then processing efficiency is improved, but noise reduction effect is not obtained for pixels subjected to defect correction
Solution Approach 1:
The patent implements a dynamic processing approach where the noise reduction unit adaptively adjusts its operation based on defect determination results. Rather than rigid parallel processing, the system dynamically routes pixel data: defective pixels bypass noise reduction, while non-defective pixels undergo noise reduction processing. This dynamic adaptation ensures both processing efficiency and effective noise reduction where applicable.
Solution Approach 2:
The patent applies noise reduction selectively to specific regions (non-defective pixels) rather than uniformly to all pixels. This local quality approach ensures that noise reduction effects are applied only where needed and where they will be effective, while defective pixels are handled separately through correction mechanisms, optimizing both efficiency and effectiveness.
3Measurement precision
If high analog gain is set due to low illuminance, then signal sensitivity is improved, but noise is intensified and mistakenly determined as defects, weakening noise reduction effect
Solution Approach 1:
The patent changes the parameter used for defect determination from absolute signal intensity to contrast-based metrics. By evaluating the difference between pixel values rather than absolute values, the system becomes insensitive to the overall noise level intensified by high analog gain. This parameter transformation allows accurate defect detection even in high-noise, low-illuminance conditions while preserving the sensitivity benefits of high gain.
Solution Approach 2:
The patent inverts the conventional defect detection approach: instead of identifying defects by their absolute signal values, it identifies them by their deviation from surrounding pixels (contrast). This inversion of the detection paradigm allows the system to distinguish true defects from noise amplification, maintaining signal sensitivity while filtering out noise-induced false defect detections.
Data Source
AI summary
According to one embodiment, an image processing device includes a defect correcting unit, a noise-reduction processing unit, and a selecting unit. The defect correcting unit executes defect correction on a target pixel. The defect correcting unit switches, according to the level of contrast determined concerning a plurality of peripheral pixels, a first correction value obtained through averaging processing for signal values of the peripheral pixels and a second correction value other than the first correction value.