Choosing The Best Barcode Recognition Sdk For A Document Imaging Processing System

By Author: Dan Manheim
Total Articles: 1
Comment this article

When building a document image processing system, a primary goal is to retrieve all critical data from scanned images with maximum reliability and accuracy. Examples of critical data are account numbers, order numbers, and customer ids which help route the document through the system. Encoding the data with barcodes and processing the barcodes with a robust barcode recognition solution provides the optimal approach for an error-free workflow.



Choosing the best solution for reading barcodes is extremely critical, since every error when reading the barcodes could have significant costs for the business. Each incorrect or failed read will cost significant time and labor to find the failed page, rescan or manually enter the failed data, and reorganize archives. The labor cost is significantly increased due to impairment that may occur with barcode images including: skew, noise, smudges, darkness, and many other barcode impairments. The goal is to find a barcode solution that will allow unsupervised operation of the production ...
... system to minimize labor costs associated with these types of errors.



To avoid these costs and find the best possible solution, considering the following factors when choosing your barcode sdk solution:

Reliability and Accuracy - When scanning 1000s of images, it's critical to strive for 100% recognition rate to avoid the costs associated with barcode recognition failure. This is particularly relevant for damaged and poorly scanned barcodes which require sophisticated algorithms for correct decoding.
Performance - Barcode recognition, like any image processing, is a computationally intensify operation. Processing thousands of documents a day requires optimized speed to ensure an uninterrupted workflow. Your barcode sdk solution should be optimized to meet the speed demands of your application.
Technical Requirements - The solution must fit with the technical requirements of the production environment such 32-bit vs 64-bit, .Net, COM / ActiveX, and localization support. The SDK should have the ability to handle your specific document types (e.g. poor quality faxed images or barcodes in PDF files)
Ease of Integration - The SDK should have a straightforward API that does not require a deep understanding of imaging and barcode recognition. Good samples and documentation should make integration straightforward.
Technical Support - You will almost certainly have questions related to barcode recognition and integration during both development and production. You should expect the highest level of expertise from your barcode recognition provider with access to the development team rather than non-technical support staff.
Cost - It's tempting to go with an open source or free solution such as Zxing, but most of these solutions focus on mobile phone applications and do not have the features necessary to support an automated workflow as discussed here. It's important to look at the total cost of ownership; this includes the cost of the development license, runtime license for your environment (e.g. Windows Service, Web Server, or unsupervised batch application), cost of production cycle interruption due to non-responsive support, and costs related to barcode recognition failures.

Doing thorough research while selecting the right barcode recognition SDK will save significant time and cost down the line when the production system needs to handle real barcodes you encounter through the lifetime of your imaging system (e.g. problematic or impaired barcodes). Find the most challenging images now, test them with an online barcode reader or by downloading a trial of a Barcode Recognition SDK to make sure the barcode sdk will handle your worst cases.