An Integration Model Based on Non-classical Receptive Fields

In this paper, we improve a new model of retinal ganglion cell receptive fields (RFs), which is different from conventional model of classical receptive fields (CRFs), and take into account the characteristic that the sizes of receptive fields vary with eccentricities. According to the information processing mechanism of retinal ganglion cell RFs including the non-classical receptive fields (nCRFs), we use the model to integrate image information. The results show that the model can not only enhance the edge contrast, but also transfer the low spatial frequencies which are filtered by the center/surround mechanism of the classical receptive field. So the image processing results are benefit for the integration of image information. The results are consistent with the spatial summation characteristic of ganglion cell RFs.