US 5,333,209 A · Method of recognizing handwritten symbols

Abstract

An unknown handwritten symbol written on a digitizing tablet is compared with symbols in a predefined "alphabet" or library of model symbols and the closest match chosen. Recognition is independent of the size, position or orientation of the symbols. The alphabet can be any collection of symbols, such as alphanumeric characters, ideograms or words in cursive script and is created by writing at least one example of each symbol on the tablet. A sequence of samples of the pen position is recorded while a symbol is being written. The samples form a vector, which is then translated so that the centroid of the symbol lies at an origin. The comparison, which can easily be done in real time, involves calculating a correlation factor from scalar products of the vector for the unknown symbol and two versions of the vector for each model symbol and choosing the model symbol having the highest correlation factor. If needed to distinguish between model symbols with similar correlation factors, the comparison can also include calculating a rotation factor from such vectors for use in making such choice. Embodiments of the invention can be configured that are user-independent, user-dependent or that evolve from one to the other.

Claims36

Dependent claims

A machine method of recognizing an unknown symbol as one of a set of stored model symbols wherein said unknown symbol and said model symbols are each represented as an ordered set of coordinate samples, which comprises:

translating the samples for said unknown symbol and each model symbol so that the centroids of said unknown symbol and said model symbol lie at a common origin,
comparing said unknown symbol with each model symbol in turn by equalizing the number of translated samples for said unknown symbol and said model symbol and determining the correlation of said unknown symbol with said model symbol from the equalized number of said translated samples for said unknown symbol and the equalized number of said translated samples for said model symbol by calculating said correlation as a quantity indicating the closeness of a representation of all possible rotations and sizes of said unknown symbol to a representation of all possible rotations and sizes of said model symbol and
recognizing said unknown symbol as the model symbol associated with the highest correlation,
whereby said recognition is independent of the size, position or orientation of said unknown symbol with respect to said model symbols.

1 more element

A machine method of finding the closest match for an unknown symbol in a set of stored model symbols wherein said unknown symbol and said model symbols are each represented as an ordered set of coordinate samples with the centroids of said unknown symbol and said model symbols at a common origin, said samples for said unknown symbol are compared in turn with said samples for each of said model symbols, before each comparison, the number of translated samples for said unknown symbol is equalized with the number of said samples for said model symbol and during each comparison, the correlation of said unknown symbol with said model symbol is calculated from said equalized and translated samples and the model symbol associated with the highest correlation is identified as the closest match CHARACTERIZED IN THAT

said correlation calculation step further comprises:
calculating said correlation as a quantity indicating the closeness of a representation of all possible rotations and sizes of said unknown symbol to a representation of all possible rotations and sizes of said model symbol,
whereby said comparisons are independent of the size, position or orientation of said unknown symbol with respect to said model symbols.

Dependencies are inferred from the claim text.

Method of recognizing handwritten symbols

Life of this patent

Method of recognizing handwritten symbols

Life of this patent

Abstract

Claims36