Vectorization

Vectorization

Automatic vectorization is a procedure of generating vector objects from original raster images. There are two types of vectorization: recognition of raster entities and approximation of raster objects.
The vectorization algorithms of the first type search for raster image fragments, which can be represented as basic geometry such as lines, arcs, circles and hatches, and create appropriate vector objects. These algorithms recognize raster analogues of vector entities. WiseImage uses a set of independent recognition algorithms, which can be used in various combinations, thus enabling users to obtain a vector image of optimal structure.
WiseImage recognizes objects of the following types: points, lines, circles, arcs, polylines, hatches, texts and symbols. The program can be trained to recognize new texts and symbols.
When vectorizing lines, circles, arcs and polylines, the program can recognize the line style of objects, and also arrows on lines and arcs. Orthogonal alignment of lines which have a small deviation from either the horizontal or vertical axis can be forced.
The approximating algorithms are used to vectorize arbitrary raster objects. Users can also approximate raster objects by outlines if they cannot be recognized as lines, arcs or circles.
Vector objects corresponding to raster lines of various widths on different layers can be placed or different colours assigned to them.
Users can also tune recognition manually or use one of the standard sets of predefined parameters – recognition templates.