| Counterfeiting is a very common phenomenon today,and preventing counterfeiting has become a major challenge in today’s world.Various counterfeit products have given rise to numerous optical anti-counterfeiting technologies.Traditional optical anti-counterfeiting technologies include optical codes,optical labels,optical color-changing materials,etc.These technologies have been applied in various fields such as currency,bills,documents,food,medicine,cosmetics,etc.,and have saved countless economic losses and even protected our lives.However,with the advancement of science and technology and the continuous development of anti-counterfeiting technology,traditional optical anti-counterfeiting technology can no longer fully meet the needs,and various anti-counterfeiting material preparation technologies are gradually being promoted,and the mysterious optical anti-counterfeiting is gradually being broken by high-tech anti-counterfeiting methods.Therefore,people are researching new optical anti-counterfeiting technologies,such as optical anti-counterfeiting technology based on nanostructures,optical anti-counterfeiting technology based on photonic crystals,etc.,to further improve the safety and reliability of products.Among many anti-counterfeiting technologies,spectral anti-counterfeiting technology is the focus of development because of its highest anti-counterfeiting information content,and counterfeit products cannot have exactly the same spectral information as genuine products.In rare earth luminescence anti-counterfeiting technology,rare earth elements are usually combined with materials such as polymers to make various shapes of marks and labels.These marks and labels can have different spectral characteristics,so they can produce specific emission spectra under specific light sources,thereby achieving the purpose of anti-counterfeiting.In addition,the emission spectra of rare earth elements are usually very stable and not affected by the environment and time,so they have high anti-counterfeiting and persistence.In response to the above situation,I have done the following work:(1)A set of Zn WO4:x mol%Er3+(x=0.5,0.7,1.0,3.0)fluorescent powder samples were prepared by high-temperature solid-phase method to study their upconversion luminescence.We irradiated this set of samples with 980nm excitation light and observed green visible light.After finding the group of samples with the strongest emission,we added sensitizer Yb3+to this group of samples and fired them to make a group of Zn WO4:1 mol%Er3+/x mol%Yb3+(x=1,3,5,10)samples to explore the effect of rare earth ion concentration on the emission of the samples.We changed the pumping power to change the emission intensity and used the law to study the upconversion luminescence mechanism of the samples.We measured the lifetime of the samples in different bands to explore their luminescence principle,and finally studied the temperature sensing characteristics of the samples using fluorescence intensity comparison technology.(2)To further enhance the luminescence intensity of the samples,I added a third ion Gd3+to the above samples.In order to weaken the influence of ion concentration saturation,I set the concentration of Yb3+at 3 mol%and changed the concentration of Er3+to prepare a group of Zn WO4:x mol%Er3+(x=0.1,0.3,0.5,1,1.5),3 mol%Yb3+fluorescent powder samples.Among them,the group with the strongest luminescence was selected and Gd3+ions were added to obtain three-doped fluorescent powder samples Zn WO4:Er3+,Yb3+,x mol%Gd3+(x=0,5,10,20,30).Up-conversion luminescence testing was performed,and it was found that the luminescence intensity of the group with Gd3+ions added was more than 100 times higher than that without Gd3+ions.The group of luminescent samples with Gd3+ions was also tested for temperature sensitivity and was found to have high sensitivity,demonstrating excellent temperature sensing ability.Therefore,this sample can be used as an optical anti-counterfeiting material.(3)This article introduces a counterfeit image recognition system based on convolutional neural networks,which was designed using the Py Charm Community Edition platform.Compared to traditional recognition methods,convolutional neural networks demonstrate significant advantages in solving problems such as slow image recognition and low accuracy.Therefore,the system can detect and identify counterfeit images more quickly and accurately,providing consumers with better protection and trust.In order to further optimize the system performance,the researchers used gradient descent to optimize the recognition system.Experimental results showed that the accuracy of the recognition system reached an amazing 99.7%.These research findings indicate that convolutional neural networks have broad application prospects in the field of anti-counterfeiting image recognition and provide important reference significance for the technological development of the anti-counterfeiting industry. |