Image Sensor Defect Detection for Adaptive Static Pixels
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing image sensors face challenges in accurately detecting defective pixels, which can occur during manufacturing or due to pixel failure or deterioration, affecting image data quality.
Innovation Solution
An image sensor with a defect detector that adaptively identifies static defective pixels in fixed locations by analyzing multiple images based on average pixel values, gain values, and light exposure times, and corrects these pixels using neighboring pixel values.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If a wafer test is performed to detect defective pixels during manufacturing, then defective pixels can be identified early, but defective pixels may still occur after manufacturing due to pixel failure or deterioration
Solution Approach 1:
The system performs preliminary defective pixel detection during wafer test using a first defective pixel map, then continuously monitors for additional defective pixels during operation by comparing current pixel values against this pre-established map. This preliminary action allows early identification of manufacturing defects while preparing for future operational failures.
Solution Approach 2:
The system implements continuous feedback by repeatedly capturing images, detecting defective pixels in real-time, updating the defective pixel map with newly detected pixels, and using this updated information to correct subsequent images. This closed-loop feedback ensures ongoing detection of both manufacturing and operational defective pixels.
2Measurement precision
If multiple images are captured and analyzed to detect defective pixels, then detection accuracy improves, but processing time and computational complexity increase
Solution Approach 1:
The system performs preliminary analysis by capturing multiple test images during initialization and generating a first defective pixel map before normal operation begins. This preliminary action consolidates the computationally intensive analysis work upfront, allowing faster real-time detection during subsequent image capture using the pre-computed reference map.
Solution Approach 2:
Instead of analyzing entire images globally for defective pixels, the system focuses computational resources on local pixel-level comparisons against the defective pixel map. By concentrating analysis on specific pixel locations identified as potentially defective rather than processing all pixels uniformly, the system achieves high detection accuracy with reduced overall processing time.
3Manufacturing precision
If defective pixels are corrected using neighboring pixel values, then image quality improves, but the correction process adds computational overhead
Solution Approach 1:
The correction process applies local quality by using only neighboring pixel values surrounding each defective pixel location, rather than processing the entire image or using complex global correction algorithms. This localized approach maintains high image quality while minimizing computational overhead by limiting the correction scope to immediate neighborhoods of defective pixels.
Solution Approach 2:
The system performs self-service correction by automatically detecting defective pixels and correcting them using available neighboring pixel data without requiring external intervention or complex external processing systems. The image sensor itself generates and applies the correction, reducing overall system complexity while maintaining image quality.
Data Source
AI summary
An image sensor includes a pixel array including pixels and a controller configured to control the pixels to produce a plurality of images. The image sensor also includes a defect detector configured to determine a target image based on an average green pixel value and a gain value related to a light exposure time among the images, and detect an adaptive static defective pixel, which is a defective pixel with a fixed location based on pixel values included in the target image.