论文信息 - A Regularized Minimum Cross-entropy Algorithm on Mixture of Experts for Curve Detection

A Regularized Minimum Cross-entropy Algorithm on Mixture of Experts for Curve Detection

Curve detection is a basic problem in image processing and remains a difficult problem. In this paper, with the help of regularization theory, we aim to solve this problem via a gradient regularized minimum cross-entropy (RMCE) algorithm on the mixture of experts (ME) model, which can automatically make model selection. It is demonstrated by the simulation and image experiments that this gradient algorithm can not only detect curves (straight lines or circles) against noise, but also automatically determine the number of curves during parameter learning

Zhiwu Lu | Zhiwu Lu

[1] Lei Xu,et al. RBF nets, mixture experts, and Bayesian Ying-Yang learning , 1998, Neurocomputing.

[2] Robert C. Bolles,et al. Random sample consensus: a paradigm for model fitting with applications to image analysis and automated cartography , 1981, CACM.

[3] Geoffrey E. Hinton,et al. Adaptive Mixtures of Local Experts , 1991, Neural Computation.

[4] Zhiwu Lu,et al. A Gradient BYY Harmony Learning Algorithm on Mixture of Experts for Curve Detection , 2005, IDEAL.

[5] Nikhil R. Pal,et al. On minimum cross-entropy thresholding , 1996, Pattern Recognit..

[6] Erkki Oja,et al. A new curve detection method: Randomized Hough transform (RHT) , 1990, Pattern Recognit. Lett..

[7] Lei Xu,et al. BYY harmony learning, structural RPCL, and topological self-organizing on mixture models , 2002, Neural Networks.

[8] Vladimir Vapnik,et al. An overview of statistical learning theory , 1999, IEEE Trans. Neural Networks.

[9] Dana H. Ballard,et al. Generalizing the Hough transform to detect arbitrary shapes , 1981, Pattern Recognit..

[10] Clark F. Olson,et al. Constrained Hough Transforms for Curve Detection , 1999, Comput. Vis. Image Underst..