A New Kind of Super-Resolution Reconstruction Algorithm Based on the ICM and the Constrained Cubic Spline Interpolation

Super-resolution reconstruction of image is highly dependent on the data outliers. This work addresses the super-resolution reconstruction design of the intersecting cortical model (ICM) algorithm applied to the constrained cubic spline interpolation. Based on a simplification of the pulse-coupled neural network (PCNN), we propose a design strategy to reduce the effects of outliers on the reconstructed image. Intersecting cortical model (ICM) has gained widely research as a new artificial neural network. It derives directly from the studies of the small mammal´s visual cortex. Cubic spline interpolation is a useful technique to interpolate between known data points due to its stable and smooth characteristics. Unfortunately it does not prevent the high-frequency information, which is essential for many image processing applications. This article presents a new interpolation method that combines the smooth curve characteristics of spline interpolation, with the non-smoothing behaviour of linear interpolation. The theory analysis and the simulation experiments of the image processing indicate that this kind of super-resolution reconstruction algorithm can not only reduce the effects of outliers effectively but also keep the details of the image sufficiently.