2D image reconstruction using bicubic interpolation

Abstract

2D image reconstruction is an important task in many applications of computer vision. The objective of image reconstruction is to recover a degraded image based on mathematical and statistical models. When an image is sent via network, image’s size is reduced to fast and low cost. Although the small size images are very useful in the data storage and image transmission, the detail and clear information can’t be received from small image. So, small image is needed to enlarge to get visual contents clearly. In enlarged image, some pixel positions are lost and needed to fill. Interpolation is the process which generates pixels to fill in the blanks. In this paper, image reconstruction approach is developed based on bicubic interpolation method. This method can produce good image with fine detailed information. This approach can be used to assist in photo editing processes and other computer vision applications.