A recent study shows that facial recognition through machine vision is technically possible. This raises interesting prospects for future machine vision applications.
Dr Guoying Zhao and Prof. Matti Pietikไinen from the Machine Vision Group of the University of Oulu have produced significant results in their research focusing on facial expression analysis through video images. They have developed two new descriptors for dynamic texture recognition and applied them to facial expression analysis.
Their article “Dynamic Texture Recognition Using Local Binary Patterns with an Application to Facial Expressions” was recently accepted for publication in the journal IEEE Transactions on Pattern Analysis and Machine Intelligence, one of the most-cited journals in electrical engineering and computer science. The journal ranks among the top journals in these fields.
The developed approach can be used in automatic and real-time facial expression recognition, and it has many potential computer vision applications. These include advanced human-computer interaction, biometric recognition, telecommunications and psychological research. For example, a machine able to recognise human emotions may soon be reality.
The approach combines motion and appearance data in a way that avoids interference from illumination variations and image transformations such as rotation. The method is theoretically and technically simple to implement.
Cite This Page: