An encoder-decoder architecture within a classical signal processing framework for real-time barcode segmentation.
- Óscar Gómez-Cárdenes (University of La Laguna)
- José Gil Marichal-Hernández (University of La Laguna)
- Jung-Young Son (Konyang University)
- Rafael Pérez Jiménez (IDeTIC, University of Las Palmas de Gran Canaria)
- José Manuel Rodríguez-Ramos (Wooptix S.L., University of La Laguna)
If using this work, please cite us.
Gómez-Cárdenes, Ó.; Marichal-Hernández, J.G.; Son, J.-Y.; Pérez Jiménez, R.; Rodríguez-Ramos, J.M. An Encoder–Decoder Architecture within a Classical Signal-Processing Framework for Real-Time Barcode Segmentation. Sensors 2023, 23, 6109. https://doi.org/10.3390/s23136109
This code accompanies the paper "An encoder-decoder architecture within a classical signal processing framework for real-time barcode segmentation".
Link to the paper: https://www.mdpi.com/2371152
The four algorithms analyzed in this paper are provided in the code:
- PDRT 2: Partial Discrete Radon Transform with
tile_size=2
. - PDRT 32: Partial Discrete Radon Transform with
tile_size=32
. - PS DRT: Partial Strided DRT based detector with
tile_size=32
andstride=2
. - MDD DRT: Multiscale Domain Detector based on the DRT.
For each of these algorithms, two different implementations are provided,
- Python with Numba: used for prototyping and design (located in
./python
) - Halide in C++: used for performance targeting different architectures (located in
./cpp
)
See python/README.md
See cpp/README.md
See license file