Status recognition technology based on extraction of facial feature information

• Published in: 2009 4th International Conference on Computer Science & Education pp. 701 - 706
• Main Authors: Yu Guofang, Li Jian
• Format: Conference Proceeding
• Language: English
• Published: IEEE 01.07.2009
• Subjects: Character recognition; contour extraction; Data mining; Ear; Face recognition; Facial features; feature comparison; Image coding; Image databases; Image processing; Image recognition; Spatial databases; status recognition
• ISBN: 142443520X; 9781424435203
• DOI: 10.1109/ICCSE.2009.5228119

A character status recognition technology is studied based on the extraction of facial feature information, i.e., based on the image processing, recognition of person face, and the locating of pupil center, the essential feature information about his/her ears, mouth, nose and face outline are extracted. By the use of double group and trip encoding, a compressed chain code with high decoding-efficiency is proposed, and the extracted feature information is encoded for comparison and the database accessed by which. In addition, combined with the parameters of matching degree and credible degree, a matching and dynamic planning algorithm is studied to implement the comparison of compressed chain codes. Finally, a comprehensive status confirming and querying experiment integrated with database, image processing and pattern recognition, as well as information comparing is carried out, the results indicate that, whether based on a image sample or a sketch sample, the system with high accuracy and efficiency of comparison and identification can be obtained as long as the image size is large enough.