Fractal-based classification of natural textures

Texture classification is an important first step in image segmentation and image recognition. The classification algorithm must be able to overcome distortions, such as scale, aspect and rotation changes in the input texture. In this paper, a new fractal model for texture classification is presented. The model is based on fractional Brownian motion (FBM). It is also shown that this model is invariant to changes in incident light; empirical results are also given. The isotropic nature of Brownian motion is particularly useful for outdoor applications, where the viewing direction may change. Classification results of this model are presented; comparisons with other texture measurement models indicate that the incremental FBM (IFBM) model has better performance for the samples tested     

Projection learning for self-organizing neural networks

A new learning scheme, called projection learning (PL), for self-organizing neural networks is presented. By iteratively subtracting out the projection of the “twinning” neuron onto the null space of the input vector, the neuron is made more similar to the input. By subtracting the projection onto the null space as opposed to making the weight vector directly aligned to the input, we attempt to reduce the bias of the weight vectors. This reduced bias will improve the generalizing abilities of the network. Such a feature is important in problems where the in-class variance is very high, such as, traffic sign recognition problems. Comparisons of PL with standard Kohonen learning indicate that projection learning is faster. Projection learning is implemented on a new self-organizing neural network model called the reconfigurable neural network (RNN). The RNN is designed to incorporate new patterns online without retraining the network. The RNN is used to recognize traffic signs for a mobile robot navigation system