Automatic reading of handwritten chemical formulas from a structural representation of the image

This paper presents a new strategy for localization and recognition of graphical entities in handwritten chemical formulas. Our system includes a first phase of global perception of the document followed by a phase of incremental extraction of the graphical entities. The phase of global perception constructs a structural representation of the drawing and provides a precise description of all the shapes present in the initial image. Thereafter this representation constitutes the main resource that will be used by different processes achieving the interpretation of the drawing. The knowledge extracted from the representation (a structural graph) is used instead of the bitmap image material to drive the interpretation process