Texture-based segmentation of natural images using neural networks

The authors describe an image segmentation method based on scalar texture measures. A neural net approach for image segmentation based on scalar texture measures is discussed. The generalization of the network approach to subsequent images in the sequence is examined. It is shown that a cascade of neural networks, where each neural network is trained on a single scalar texture measure, can be used for image segmentation. Experimentation has been used to demonstrate the feasibility of this technique