Image texture line vectorization system and method

Abstract

The invention discloses an image texture line vectorization system and method. The method comprises the following steps: firstly, a single-pixel line skeleton of a texture image to be vectorized is extracted; secondly, the obtained line skeleton diagram is input into a preset neural network model, and the line skeleton diagram is decomposed into a plurality of smooth intersection-free paths afterpassing through a neural network; then, according to the paths, corresponding area lines are found in the original image, if the corresponding areas are equal-width lines, Bezier curve fitting is used, a framework is expanded directly according to the line width and color information, and the framework is written into an EPS format file; if the corresponding area is a non-equal-width line, the closed contour path of the area needs to be extracted, Bezier curve fitting is used, the closed path is filled with a corresponding color, and the color is written into an EPS format file; finally, the vectorization results of all the areas are combined and displayed to a user, the user can finely adjust the contour on the vectorization results, and an EPS format vector diagram is output.

Description

technical field [0001] The present invention relates to the field of image processing, in particular to a system and method for vectorizing image texture lines. Background technique [0002] The more common forms of digital images are bitmaps (raster images) and vector graphics. Bitmap has better color display effect, but its display effect depends on the resolution of the image. Vector graphics draw graphics based on geometric properties, and there will be no jagged effect after zooming in. But the color display of vector graphics is relatively not so rich. At present, the more common image vectorization algorithms are: polygon-based bitmap outline vectorization algorithm, subdivision-based image vectorization algorithm. The former is to decompose the original image into multiple closed parts, and perform path tracing and color filling on these paths respectively to complete vectorization. Its advantage is that the amount of calculation is small, but the vectorization e...

AI Technical Summary

Problems solved by technology

Its advantage is that the amount of calculation is small, but the vectorization effect for complex texture images depends on the result of path tracing, and it cannot vectorize images with complex colors well.

The latter uses subdivided surface patch vertices to record information such as graphic color, position, and color change for vectorization. Its advantage is that it can vectorize complex color image blocks such as gradient colors, but the vectorization effect for texture lines is general.

Images