大人群指掌纹高精度识别技术及应用

清华大学自动化系研究组在图像处理、特征表达、特征提取、图像比对等方面开展了长达十余年的研究项目，实现了大人群下高精度的指掌纹识别，解决了四大难题：

1.通常的指纹图像质量参差不齐，比如模糊、重叠、残缺等；

2.人的手指是立体柔软的而图像是平面的，很容易出现扭曲；

3.1对1的指纹识别是较为简单的，而1仅N的识别就相对困难；

4.遇到只有掌纹而没有指纹的情况，直接套用指纹识别技术效果不理想。

Technology of High-Precision Fingerprint and PalmprintRecognition in A Large Population and Its Applications (China)

The research group of the Department of Automation of Tsinghua University has carried out more than ten years of research projects in image processing, feature expression, feature extraction, image comparison, etc., achieving high-precision finger and palmprint recognition in a large population and solving four major problems:

1. Usually the quality of fingerprint images is uneven, such as blur, overlap, incompleteness, etc.;

2. Human fingers are three-dimensional and soft but the image is flat, which is prone to distortion;

3. One-to-one fingerprint recognition is relatively simple, while one-to-N recognition is relatively difficult;

4. When there are only palm prints but no fingerprints, the result of directly applying fingerprint recognition technology is not ideal.