I tested the class for Barcode detection and decoding and whereas the detection of my EAN-13 barcodes worked well, the algorithm failed to decode some of them consecutively.
In some cases the corner points are in the wrong order but also after rearranging their order the decoding failed.
I tried to use the DNN model and downloaded the two files as recommended in the tutorial ( OpenCV: Barcode Recognition ) but the inclusion of these two files in the constructor had strictly NO effect on the results!
What is your experience with the DNN model?
Unfortunately, there are no parameters to adjust in the decoder (decodeWithType()) but the two parameters detectorScales and gradientThreshold improved the detection results.
Do you have some advice how to improve decoding?
Thank you in advance!