Wafer Defect Detection Using Radial Line Scan Cameras
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional image processing methods for defect detection on wafers require converting rectangular images captured by rotating cameras into circular images, which is inefficient and results in a larger detection device size due to the need for horizontal and vertical driving mechanisms.
Innovation Solution
A method and device that acquire and pre-process rectangular images of a wafer's inner and outer ring portions, determining pixel mapping relationships to restore a circular image, reducing the device size by using line scan cameras arranged radially and splicing cropped images to form a complete rectangular image of the wafer.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If rectangular images are captured by rotating cameras and converted to circular images, then the wafer can be detected, but the device size increases due to horizontal and vertical driving mechanisms
Solution Approach 1:
The patent divides the wafer detection into two independent parts: inner ring portion and outer ring portion, each captured by dedicated line scan cameras. This segmentation eliminates the need for complex horizontal and vertical driving mechanisms, as each camera only needs to rotate independently to scan its respective ring portion, thereby reducing device size while maintaining detection capability.
Solution Approach 2:
The patent transitions from traditional 2D area scanning to 1D line scanning in a radial dimension. By using line scan cameras arranged radially and rotating them to capture circular ring portions, the system achieves complete wafer coverage without requiring horizontal and vertical driving mechanisms, thus reducing device volume while preserving measurement precision.
2Measurement precision
If rectangular images are captured and restored to circular images, then the wafer can be observed, but the processing efficiency decreases
Solution Approach 1:
The patent segments the wafer into inner and outer ring portions, each captured by dedicated line scan cameras. This segmentation allows for parallel processing of image data from both rings simultaneously, significantly improving processing efficiency while maintaining the ability to observe the complete wafer structure.
Solution Approach 2:
The patent performs preliminary processing by capturing and pre-processing rectangular images of inner and outer ring portions separately before final assembly. This preliminary action enables optimized processing of each ring portion independently, improving overall processing efficiency while preserving complete wafer observation capability.
3Volume of stationary object
If line scan cameras are arranged radially to capture inner and outer ring portions, then device size is reduced, but image splicing complexity increases
Solution Approach 1:
The patent segments the imaging process into two independent radial scans (inner and outer rings), each handled by dedicated line scan cameras. This segmentation simplifies the overall system architecture by eliminating the need for complex horizontal and vertical driving mechanisms, and the image splicing is managed through parameter-based coordination of the two independent scans.
Solution Approach 2:
The patent uses parameter changes (such as redundant width, overlapping width, quantity of rows) to coordinate and align the inner and outer ring portion images during splicing. By managing complexity through well-defined parameters rather than complex mechanisms, the system achieves reduced device size while keeping processing complexity manageable.
Data Source
AI summary
The present application discloses an image processing method and apparatus, an optical system, and a computer-readable storage medium. The method comprises: acquiring parameters of rectangular images of inner ring portions and outer ring portions of an object to be inspected; preprocessing, according to the parameters of the rectangular images of the inner ring portions and the outer ring portions, the rectangular images of inner ring portions and the rectangular images of outer ring portions to form a rectangular image of the object; initializing parameters of a target circular image according to the rectangular image of the object; and determining pixel mapping relationships between the target circular image and the rectangular image of the object, and determining pixel values on the target circular image according to the pixel mapping relationships and pixel values on the rectangular image of the object.