Connected and Overlapping Symbols

When symbols are connected or overlap, there is the problem of separating them. This is more of a problem with scanned input, as the input is simply an image. For handwritten input, there is information available on the timing and order of the strokes drawn.

  

Figure 2.4: Overlapping symbols.

Figure 2.4 shows another problem, where the tail of the y overlaps the fraction bar and the x. This makes it hard to determine the geometric relationship between these symbols reliably. Anderson  initially represents the position of each symbol with a rectangular bounding box, the smallest rectangle that contains all of the symbol's original pixels. He then shrinks these bounding boxes to a single centre point to avoid problems when considering the relationships between symbols. The position of the centre point is dependent on the identity of the character. The way Anderson defines the centre points of bounding boxes is covered in more detail in the discussion of his equation parser in Section .