TFT Sensor Image Processing for CCD Algorithm Compatibility
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Fingerprint capture systems using thin-film transistor (TFT) sensors produce images with inverted gray levels and cast shadows, making it difficult for algorithms designed for CCD or CMOS sensors to process them effectively, as these algorithms are unsuitable for images with different properties.
Innovation Solution
A method involving image processing steps such as high-pass filtering, averaging filtering, connected component analysis, hierarchical tree organization, convex hull determination, and pixel value modification to transform TFT sensor images into properties compatible with CCD or CMOS sensor images, allowing for the use of TFT sensor images with algorithms intended for CCD or CMOS sensors.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If TFT sensor is used to capture fingerprint images, then the imaging system can be implemented with TFT technology, but the resulting images have inverted gray levels and cast shadows that make them incompatible with algorithms designed for CCD or CMOS sensors
Solution Approach 1:
The patent introduces an image processing module as an intermediary between the TFT sensor and the fingerprint recognition algorithm. This module transforms the TFT sensor output (with inverted gray levels and shadows) into a format compatible with algorithms designed for CCD/CMOS sensors, allowing the use of proven algorithms without modifying them while enabling TFT sensor usage
Solution Approach 2:
The patent applies parameter transformation to the image data by inverting the gray levels and removing shadow components through image processing operations. This changes the parameters of the image (brightness distribution, contrast) to match the expected input characteristics of existing fingerprint recognition algorithms
2Adaptability or versatility
If image processing steps are applied to transform TFT images, then compatibility with CCD/CMOS algorithms is achieved, but the processing complexity increases
Solution Approach 1:
The patent performs preliminary image processing transformations (gray level inversion, shadow removal) on the TFT sensor output before it enters the fingerprint recognition algorithm. By preparing the image in advance to match the expected input format, the existing algorithms can be used without modification, and the processing complexity is confined to a dedicated preprocessing module
3Ease of operation
If standard fingerprint processing algorithms are used directly on TFT images, then the processing is simple, but the algorithms cannot effectively process images with inverted gray levels and shadows
Solution Approach 1:
The patent inserts an image processing intermediary that transforms TFT sensor images into a format suitable for standard algorithms. This allows the remainder of the processing pipeline to use simple, well-established fingerprint recognition algorithms without modification, maintaining processing simplicity while ensuring effectiveness through the transformation layer
Data Source
Figure 1~2
Figure 3~4
Figure 5A~7
AI summary
A method for processing a representative image of at least one finger provided by a thin-film transistor image sensor of a fingerprint capture system. The method aims to modify characteristics of said image so that they correspond to the characteristics of images acquired by a fingerprint capture system using a CCD or CMOS sensor, while preserving relevant image information.