Abnormal Pixel Detection in Image Sensors Using Multi-Stage Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
It is challenging to accurately distinguish between normal and abnormal pixels in an image sensor, especially when there are a large number of abnormal pixels, as peripheral pixels may also be defective, leading to deteriorated detection accuracy.
Innovation Solution
An image processing device employs multiple specification methods to identify abnormal pixels, using different criteria to enhance detection accuracy, including statistical methods based on frequency distributions and noise ratios, allowing for stepwise specification and correction of abnormal pixel outputs.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a single detection method is used to identify abnormal pixels, then the detection process is simple and fast, but the detection accuracy deteriorates when abnormal pixels are numerous
Solution Approach 1:
The detection process is divided into multiple sequential stages: first detection unit identifies abnormal pixels using a first detection method, then a second detection unit identifies remaining abnormal pixels using a second detection method. This segmentation allows each method to operate on a reduced subset of pixels, maintaining accuracy while managing complexity.
Solution Approach 2:
The patent applies partial action by using the first detection method to identify and remove a portion of abnormal pixels, then applying the second detection method only to the remaining pixels. This excessive action ensures that even if one method fails to detect all abnormal pixels, the second method provides additional detection coverage.
2Productivity
If peripheral pixels are used for detection, then the detection process is efficient, but detection accuracy deteriorates when peripheral pixels are also abnormal
Solution Approach 1:
The first detection unit performs preliminary detection to identify and remove abnormal pixels before the second detection unit operates. This preliminary action ensures that when the second detection unit uses peripheral pixels for detection, those peripheral pixels are more likely to be normal, thus maintaining both efficiency and accuracy.
3Measurement precision
If multiple detection methods are applied to all pixels, then detection accuracy improves, but processing time and computational load increase significantly
Solution Approach 1:
The pixel set is segmented into different groups that are processed by different detection methods. The first detection method processes all pixels, then the second detection method processes only the remaining pixels. This segmentation reduces the total computational load compared to applying both methods to all pixels.
Solution Approach 2:
The second detection method is applied partially only to pixels that were not identified as abnormal by the first method, rather than applying it excessively to all pixels. This partial application reduces processing time while maintaining the benefit of multiple detection methods.
Data Source
AI summary
In an aspect, an image processing device includes: first specification means for specifying an abnormal pixel from a plurality of pixels according to a first method; and second specification means for specifying an abnormal pixel, from the plurality of pixels excluding the abnormal pixel specified by the first specification means, according to a second method different from the first method.