The situation changed when camera equipped devices (like mobile phones) and document scanners became common for everyday usage.
Implementation of a barcode writing library does not require hard Computer Science skills because it just need to follow AIM[13] or ISO specifications.
[15] These differences could not be solved by barcode fonts usage and required API with multiple parameters processing.
This mode captured the whole image but then library made some scan-lines with Bresenham's algorithm and tried to recognize data from these lines as hardware laser scanners did.
The bright representation of these libraries is early ZXing project supported by Google, ZBar[16] or other solutions.
Barcode libraries have provided low cost automatic identification and data capture features to various fields of services and industry.
This is solved with predefined different metadata values in set of fonts for the same type of barcode.
Barcode libraries with API calls have more customization features in writing and reading modes.
All other ways like RFID chips or object detection by image recognition are more expensive and difficult to implement.
Main advantages of 2D barcodes are high encoding density, which is 10 times more, no restrictions to text encoding and self-checked codes like Reed Solomon codes, which not only add confidence in correct recognition but also can restore some wiped or corrupted barcode data.
[86][87] They do not have any restrictions to encoding text, they can be correctly restored on corrupted or low quality images and their recognition result is fully confidential.