A Locally Orientation-Adaptive And Fusion-based Algorithm For Image Interpolation

Image information is the most important way to get information. With the rapid development of information technology, especially the improvement of the requirement to image quality, the image technic deeply affects all aspects of our life. Image interpolation technology has been an active technology for decades.At present, there has been a lot of classical and popular algorithms, in which the traditional linear interpolation algorithm is the most widely used one in all fields for its simple calculation. However, the results are not satisfactory. In this thesis, basised on the existing classical algorithms, a locally orientation-adaptive and fusion-based algorithm is proposed. The algorithm, making use of the local image gradient information and data fusion technology, got the optimized estimated point. The algorithm is applied to gray-scale image to enlarge the image and to color mosaicing image to reconstruct the full color image. The calculation of this algorithm is simple, and the edges of the image reconstructed are smooth. The experimental results showed that the method proposed in this thesis outperformed the former methods espesally in visual perception.