Robert Adelmann An efficient bar code recognition engine for enabling mobile services. PhD thesis No. 19721, ETH Zurich, Zurich, Switzerland, 2011 Abstract: In the area of pervasive computing, mobile phones have evolved into attractive development platforms that show considerable potential when it comes to bridging the often-cited gap between the real and virtual world. They are ubiquitous, highly mobile, provide significant computing power, and increasingly also offer an abundance of built-in sensors. With the general availability of smartphones and affordable data rates, consumers are beginning to use their mobile phones to interact with physical products found in stores in order to access product-related information and services. To support this interaction, consumer-oriented mobile applications require a fast and convenient way to identify products. Even though Near Field Communication (NFC) and Radio-Frequency Identification (RFID) technology is very promising for that purpose, the widespread use of RFID tags on retail products remains unlikely for the next years. In contrast, bar codes are ubiquitous on virtually all packaged consumer goods world-wide. Recognizing bar codes with mobile phones induces many challenges, however. Blurry images from cameras without autofocus, but also shadows and glare, low video-image resolutions and limited computing power on many mass-market phones are just some of the inherent difficulties. Furthermore, developing applications for mobile phones still requires considerable expertise, despite the fact that mobile phones have evolved into attractive development platforms over the past years. This thesis addresses these issues with two main contributions: First, it presents the design and implementation of a bar code recognition method for mobile phones that addresses the aforementioned challenges and, in particular, is capable of recognizing bar codes in blurry images. It provides a comparison with existing mobile bar code recognition engines on mobile phones and shows that the method presented outperforms other solutions in terms of scan speed and accuracy. The second contribution consists of a rapid prototyping environment for mobile phones that enables even non-experts to develop novel mobile services that leverage the bar code recognition in a fast and easy way.