The Automatic Writer Identification System on Computer (AWISC) developed by Tsinghua University passed the technical appraisal organized by the Ministry of Public Security in China on Jan. 10, 2005. Designed and implemented by Prof. Xiaoqing Ding's research group in Tsinghua University, under the cooperation with the Second Research Institute of the Ministry of Public Security in China (SRIMPS), the system realized the automatic function of writer identification on computer and can efficiently identify whether two Chinese handwriting samples are from the same individual or not. As the system also has following functions: automatic generalization of feature comparison map through segmentation and recognition of handwriting samples using Optical Character Recognition (OCR) technique, automatic generalization of identity identification book, it reduces the graphology experts' work greatly and makes the identification result more objective than manual identification in traditional way.

The appraisal conference was held in SRIMPS on Jan. 10, 2005. The general appraisal committee chair is Ruwei Dai, member of Chinese Academy of Sciences. The members of the committee include experts from the area of writer identification and trace examination. Leaders of the Ministry of Public Security in China also attend the meeting.

The Experts gave high praises to the system. The final appraising decision is that AWISC solves the automatic Chinese writer identification problem for the first time in the world, and the system performance is at the world leading level.

A person's handwriting has his/her own writing characteristic. Writer identification is the process of identifying the writer by comparing some attributes of the test handwriting sample with the reference handwriting samples. Writer identification can be served as evidence of deciding possible criminals and testimony in forensic justice.

According to the experts, traditionally writer identification is fulfilled by specialist, and the result tends to be subjective. Furthermore, the tasks are heavy and tedious. While AWISC can overcomes these problems.

The successful implementation of Automatic Writer Identification System on Computer not only is a great progress in biometric feature identification research area in China , but also provides a powerful tool for the police.

Reported by Liangrui Peng and Xianliang Wang, from Dept. of Electronic Engineering