Image Sensor Testing via Frequency Domain Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional methods for testing CMOS image sensors are inadequate in detecting defects, particularly unusual noises that are not detected by the standard mobile imaging architecture (SMIA) standard, as the complexity of CMOS image sensors increases.
Innovation Solution
A method and system utilizing frequency domain processing to test image sensors by capturing test images, generating composite images, and analyzing frequency data to determine defects, including band pattern noise that is undetectable by conventional methods.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional SMIA standard testing methods are used, then the testing process is simple and fast, but unusual noises and defects in complex CMOS image sensors cannot be detected
Solution Approach 1:
The patent transforms the test image from spatial domain to frequency domain using Fourier transform, enabling detection of unusual noises and band pattern defects that are invisible in the spatial domain. This dimensional transformation allows the testing method to detect complex defects in advanced CMOS image sensors while maintaining systematic analysis.
Solution Approach 2:
The patent changes the analysis parameters by converting images to frequency domain and extracting specific frequency components. By analyzing frequency spectra and identifying abnormal frequency patterns, the method detects defects that conventional spatial domain methods miss, improving measurement precision without requiring complex hardware modifications.
2Measurement precision
If frequency domain processing is applied to detect unusual noises, then defect detection accuracy improves, but processing time and computational complexity increase
Solution Approach 1:
The patent extracts only the essential frequency components and defect-related features from the full frequency spectrum, rather than processing the entire image data set. By focusing on specific frequency ranges and defect characteristics, the method reduces computational load while maintaining high detection accuracy.
Solution Approach 2:
The patent applies frequency domain processing selectively to identify critical defect patterns, using partial processing of frequency data where full processing would be unnecessary. This approach balances computational effort with detection accuracy, reducing processing time while maintaining effectiveness.
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
This approach allows for accurate and efficient detection of defects in image sensors, improving the reliability of defect identification and manufacturing processes.
Implementation Method 1
a plurality of test images are generated, by the image sensor that is a device under test (DUT), by capturing light generated from a polarization light source or by capturing an environment without a light source
Data Source
AI summary
In a method of testing an image sensor, at least one test image is captured using the image sensor that is a device under test (DUT). A composite image is generated based on the at least one test image. A plurality of frequency data are generated by performing frequency signal processing on the composite image. It is determined whether the image sensor is defective by analyzing the plurality of frequency data.