Read barcodes from mobile camera vs. scanned documents

Before v4.2, Dynamsoft Barcode Reader SDK has been focusing on barcode detection on scanned documents. In version 4.2, the mobile edition was introduced with some optimization implemented to better support barcode scanning from iOS and Android phones.

In this article, let’s go through differences on barcodes scanning from mobile cameras and scanned documents.

1. Image tilt and distortion
   When recognizing barcodes with a phone, it is common that the captured image is tilted on the z-axis, or distorted because the barcode is attached to an uneven surface. This requires image processing procedures, like perspective correction, to remove the distortion.Things are much easier with scanned documents, since there are only two dimensions (x-axis and y-axis) with scanning.
2. Bit depth
   Scanned images are typically black/white grey. With cameras on mobile, the images are mostly colorful. So information in the images might be removed during the “binarizing” process, and this might cause difficulties in the barcode decoding step.
3. Brightness and contrast
   With scanned documents, one images has the same brightness and contrast across the image.With cameras on mobile, brightness and contrast is uneven in one image.
4. Size of the barcode image
   With scanned documents, the barcode usually takes a small portion of the image. If the document is scanned in a low resolution, image cleanup steps will need to be performed to make the image clearer. In some cases, the barcode might even become unreadable.When reading barcodes on mobile, users hold the phone and point the camera to the barcode, the barcode takes up a big portion. In many cases, this means more pixels per inch, and thus better accuracy.
5. Image vs. Video stream
   Scanned documents are still images.With mobile phones, users adjust the phone constantly until the barcode is recognized.