Abstract:The authenticity identification of anti-counterfeiting codes based on mobile phone platforms is affected by lighting environment, photographing habits, camera resolution and other factors, resulting in poor collection quality of anti-counterfeiting codes and weak differentiation of anti-counterfeiting codes for high-quality counterfeits. Developing an anti -counterfeiting code authentication algorithm based on mobile phones is of great commercial value. Although the existing algorithms developed based on special equipment can effectively identify forged anti-counterfeiting codes, the anti-counterfeiting code identification scheme based on mobile phones is still in its infancy. To address the small differences in texture features, low response speed and excessively large deep learning models used in mobile phone anti-counterfeiting and identification scenarios, we propose a feature-guided double pool attention network (FG-DPANet) to solve the reprinting forgery problem of printing anti-counterfeiting codes. To address the slight differences in texture features in high-quality reprinted anti-counterfeiting codes, we propose a feature guidance algorithm that creatively combines the texture features and the inherent noise feature of the scanner and printer introduced in the reprinting process to identify anti-counterfeiting code authenticity. The introduction of noise features effectively makes up for the small texture difference of high-quality anti-counterfeiting codes. The double pool attention network (DPANet) is a lightweight double pool attention residual network. Under the condition of ensuring detection accuracy, DPANet can simplify the network structure as much as possible, improve the network reasoning speed, and run better on mobile devices with low computing power. We conducted a series of experiments to evaluate the FG-DPANet proposed in this paper. Experimental results show that the proposed FG-DPANet can resist high -quality and small-size anti-counterfeiting code reprint forgery. By comparing with the existing algorithm based on texture, it is shown that the proposed method has a higher authentication accuracy. Last but not least, the proposed scheme has been evaluated in the anti-counterfeiting code blurring scene, and the results show that our proposed method can well resist slight blurring of anti-counterfeiting images.

Fast recognition method for printed numbers

Digit Recognition Based on Programmable Nanophotonic Processor

Banknote Serial Number Recognition Using Deep Learning

Identification Technology Based on Geometric Features of Tooth Print Images

Large-Scale Printed Chinese Character Recognition for ID Cards Using Deep Learning and Few Samples Transfer Learning

Real-Time Recognition Of Series Seven New Zealand Banknotes

Printed Chinese Optical Character Recognition by Neural Network.

Design of Real Time Digits Recognition System for Automated Guided Vehicle

A Mask RCNN based Automatic Reading Method for Pointer Meter

Algorithm for Numeral Recognition

A Method to Locate and Recognize LCD Display Screen Using FCN

Identification of a Printed Anti-Counterfeiting Code Based on Feature Guidance Double Pool Attention Networks

A Hybrid Handwritten Digits Recognition System Based on Neural Networks and Fuzzy Logic

Printer source identification of quick response codes using residual attention network and smartphones

A New Method of the Image Pattern Recognition Based on Neural Networks.

Identifying People's Faces in Smart Banking Systems Using Artificial Neural Networks

Face Authentication System Using the Chinese Second Generation Identity Card

A Novel High-Performance Face Anti-Spoofing Detection Method

Investigation on Intelligent Recognition System of Instrument Based on Multi-step Convolution Neural Network

A Fast Face Recognition System Based on Deep Learning

Classification modeling method for hyperspectral stamp‐pad ink data based on one‐dimensional convolutional neural network