Hu & Zernike moments have always been used for grey image representation. In this study we have tried to employ them directly for color image description. This would enable us to keep the maximum amount of information given by the image colors. Regarding the classification process we have opted for the neural networks classifier, which enable to implicitly detect complex nonlinear relationships between dependent and independent variables, and to detect all possible interactions between predictor variables, and the availability of multiple training algorithms. In this document, we present a comparative study between different 3D color objects recognition systems. We have used a variety of topologies of Neural Multi-layer Networks (simple, nested and parallel networks), to come up eventually with a suggestion of a multi-Oriented Neural Networks.
展开▼
