A Neuro-Fuzzy Model for Motion Cognition

Motion tracking is a traditional problem that was tackled by various approaches including statistical methods and dynamical filtering techniques. In this paper, we introduce an approach that models this problem as a cognitive process in which motion is captured and stored as knowledge and hence can be recalled or recognized. The adaptive neuro-fuzzy inference system (ANFIS) model is used as a cognitive model in order to model and represent motion features. Motion dynamics and curvature are model inputs and the tracked object positions are the output. The model is tested on a set of motions representing maneuvering and non-maneuvering targets and it successfully tracked all motions. Moreover, the model has the ability to learn more quickly. The results are more accurate in comparison with similar work using feed forward neural network (FFNN).