Skip to Main Content
In order to reconstruct the edges of micro-image and guarantee the fitting precision of the boundaries, the wavelet transform and improved cubic spline interpolations algorithm are effectively combined. Contrapose to the complexity of the target image edges, the lifting scheme is put forward, in the processing of the wavelet transform of the image different wavelet base is chosen, the purpose of it is to reduce the interference, which is made by the high frequency random noise to the image edges. Then the general cubic spline interpolations is developed, the improved cubic spline is convex, it can restore the feature points of the image edges that were ever interfered by the noise, further more can keep the convex approximation of the image edges. Finally, high-resolution image with highlight edges is reconstructed. The experiment data indicates that the improved algorithm is better than other general algorithms, the PSNR of the reconstructed image is increased at least 1 dB.