A Study On Image Editing Based On Curve Information

Image processing is widely used in daily life, and this problem can be divided into two categories:image processing based on pixel color data and image processing based on image structures. The latter is more stable and accurate, but hard to be implemented. The edges of the image are important parts of its structure data, most of them are in the shape of curves or polygonal lines, and they can be used as the evidence of image processing after being ex-tracted.Curvature is used in traditional geometry field to measure the bending degrees of curves. But curvature is susceptible while the curve zooming and that reduce its applications greatly because it cannot judging the similarity of zoomed curves. To solve this problem, the concept of quasi-curvature is utilized in this paper, which is used to measure the similarity of curves. Quasi-curvature is similar with curvature; it can also be used to measure the bending degree of curves, but not been affected by rotating, translation or zooming. The parameter intervals of the similar sections between two different curves can be calculated at the same time. This method is used in image mosaic, image alignment and image object retrieval, and examples illustrate the effectiveness and efficiency of the new method.To use the method of quasi-curvature in image processing, the first problem to be solved is to extract the curves in the image. One of the commonly used curves in the image is the edge. To get more accurate edges, a better gray-scale image with higher contrast is needed. Since the traditional gray-scale methods are static, though the result can be adjust by some parame-ters manually, but once the parameters are confirmed, the result of every pixel is confirmed at the same time, and can’t optimized automatically acceding to the colors in the image. This paper provides a new method to convert a color image into a gray one which is based on the positions of pixies in RGB color space. The new method can use the gray-scale space suffi-ciently and improve the contrast of the resulting picture, therefore make it clear and easy to identify, and will do benefit to further process. Improved methods are also provided for better efficiency and better effect.Finally, this paper applies the two above methods in image processing such as fixing and mosaic, which can lead to a good result.