Abstract
The edge features in facial images change with changes in images, and are useful for automatically recognising the facial expressions. We present an edge feature extraction method using neuronal threshold logic models to automatically recognise the facial expressions. The system is simulated at digital system level consisting of reading an image followed up with edge extraction system that can be implemented with hybrid CMOS-memristive digital circuits. The results indicate robust edges of the facial features and its useful in development of real-time emotion recognizing digital chip.