Word Recognition for Checks in Bank Systems{4}


A database called SIDB is being developed that focuses on word recognition algorithms and is to be used with banking systems to read checks. ¬†Currently, many databases already exist but there are no free ones that banks can use to test it. According to the article, “This database in our experiments is used to support the research toward the development of recognition systems…”. Many banks that do successfully implement these systems have privacy issues wherein they simply cannot share their database for others to use. The research on this database aims to standardize word recognition in banks that want to implement systems that automatically read checks. SIDB supports basic words and numbers. For the words, they took several samples from writers of different backgrounds (different age, sex, cultural levels). In addition to the words and numbers already recognized by SIDP, they also process the image using the following: binarization, noise reduction, skew correction, slant correction, smoothing, and a control box (where the image is trimmed to only the handwritten portion).

This project is a great example of how the community respond to changing industry standards to application development. Being able to have checks that are automatically read by banking systems will soon be the de facto standard in the industry. However, the problem here is that the banks that do implement this feature are not able to share the technology with other banks simply because of security issues. And therefore a company that comes out with a free system that includes provided algorithms for accurately reading checks benefits the entire industry.

Basic knowledge of how databases work can be applied in a variety of industries. Many databases, in this example those used in banking systems, contain very sensitive material and and thus security is a very important aspect that cannot be overlooked especially when it involves people’s banking information. And thus its important that database administrators take all necessary precautions in securing databases so as to protect sensitive information gathered from their clients.

References:

Impedovo, S.; Facchini, G.; Mangini, F.M.; , “A New Cursive Basic Word Database for Bank-Check Processing Systems,” Document Analysis Systems (DAS), 2012 10th IAPR International Workshop on , vol., no., pp.450-454, 27-29 March 2012
doi: 10.1109/DAS.2012.5
URL: http://0-ieeexplore.ieee.org.opac.library.csupomona.edu/stamp/stamp.jsp?tp=&arnumber=6195412&isnumber=6195323