| Digital image plays a very important part in modern society,and it has been applied in every aspects of life.But pretty-looking images on video monitors result in low quality blurred images after magnified and printed.If hardware is improved on to increase resolution,higher expense will be paid;High resolution images also can not been stored and transmitted directly on the line because of limitation of storage space,net band broad and transmission rate.So valid method for image interpolation has become the aim which people are pursuing,and image magnification technology emerges as the times require.The main limitation of state-of-the-art magnifying techniques is that they do not introduce any new information to the original image.Some classical image interpolation magnification methods approach to perfect low pass filters.So high frequency parts are lost, and in this way the edges of images become blurred.Moreover,the edge is the one of most important characters in the image.So,many methods all pay attention to improving the edge. The idea underlying this work takes an edge-directed approach by using the sub-pixel edge map of the original low resolution image to guide the interpolation process.Sub-pixel edge location is the first step of this method.In this paper,we introduce the ROT filter and the LOG edge detection operator which can adjust the detection results by controlling the parameters and is better for the case of a narrow edge.Secondly,this paper elaborates edge-directed interpolation methods specific process and proposes a new non-oscillatory interpolation curve which can decrease the flaws generated from polynomial curve.Finally,experiments show that the method is effective.The method has been applied to several sets of different images.Respectively,from the visual evaluation and quantitative estimate,we get clear images and good numerical results.Compared to the original method, the magnified images have fewer flaws,and this method have a better anti-noise ability. |
PDF Full Text Request