Abstract: A method for exploiting correlated mail streams using optical character recognition is provided in which a static database is used to store data based on training. Real-time data for the parameters of interest, such as address block location, zip code, city, state, and font size or type is collected from the mail processing equipment in order to generate a statistical information database. The dynamic database can include probability density functions, correlations statistics, mean, variance, and high order moments. The statistical parameters are tracked using recursive least squares schemes with various windowing options, as will as moving average linear filters. Based on cost models which indicate the cost of making various types of errors in the OCR process, a decision threshold is determined which is based on the real-time statistics of the mail stream.

Abstract: A method of cursive address recognition of mail pieces using adaptive dictionary management is provided which includes creating a plurality of word databases containing the most frequently observed words in a particular portion of an address block derived from training data. Word images from various portions of the address block are extracted and the word images are compared to the database corresponding to the appropriate portion of the address block using a cursive word recognition engine. The search order of the word databases are updated based on the frequency of occurrence of recognized words in a predetermined number of previously evaluated addresses.

Abstract: A method of cursive address recognition of mail pieces using adaptive dictionary management is provided which includes creating a plurality of word databases containing the most frequently observed words in a particular portion of an address block derived from training data. Word images from various portions of the address block are extracted and the word images are compared to the database corresponding to the appropriate portion of the address block using a cursive word recognition engine. The search order of the word databases are updated based on the frequency of occurrence of recognized words in a predetermined number of previously evaluated addresses.