TWO IMPROVED MODELS OF RETINAL GANGLION CELL RECEPTIVE FIELD

摘要：

We improve two mathematical models of retinal ganglion cell receptive fields proposed by Li Chaoyi et al. and Ghosh et al. respectively, in order to highlight the mutual inhibition among the cell responses within a disinhibitory non-classical receptive field. Then we build a two-layer neural network model in which the improved mathematical models are used. When they are exposed to natural images, the improved mathematical models can not only better enhance the image edge contrast, but also decrease the image region contrast which is maintained by the nonimproved ones. The results suggest that, compared with the non-improved ones, the improved mathematical models may produce the processing results that are more conducive to the subsequent image segmentation.

关键词： retina receptive field mutual inhibition disinhibition model

作者: X.M.Wang H.Wei

作者单位: School of Computer Science,Fudan University,Shanghai,China

会议类型: 国际会议

会议名称: 2008年拟人系统国际会议(2008 International Conference on Humanized Systems )(ICHS’08)

会议地点: 北京

会议语种:英文

在线出版日期: 2008-10-18（万方平台首次上网日期，不代表论文的发表时间）