基于DeepLabv3的隨機(jī)褶皺防偽圖案識(shí)別研究
信息技術(shù)與網(wǎng)絡(luò)安全
陳 雨1,,陳桂雄1,周雄圖1,,2,,張永愛(ài)1,,2,林志賢1,,2,,吳朝興1,2,,郭太良1,,2
(1.福州大學(xué) 物理與信息工程學(xué)院,福建 福州350116,; 2.中國(guó)福建光電信息科學(xué)與技術(shù)創(chuàng)新實(shí)驗(yàn)室,,福建 福州350116)
摘要: 針對(duì)現(xiàn)有防偽技術(shù)可靠性較低、容易被仿制,、防偽成本高昂等問(wèn)題,基于DeepLabv3,,提出一種由熱膨脹系數(shù)失配產(chǎn)生壓縮應(yīng)力形成隨機(jī)褶皺防偽標(biāo)識(shí)圖案的識(shí)別方法,。具體采用深度卷積網(wǎng)絡(luò)分類(lèi)算法中DeepLabv3進(jìn)行分類(lèi)識(shí)別,通過(guò)優(yōu)化全連接層并設(shè)置不同的神經(jīng)元節(jié)點(diǎn),,提高識(shí)別網(wǎng)絡(luò)的分類(lèi)準(zhǔn)確率,,縮減訓(xùn)練時(shí)間,,訓(xùn)練準(zhǔn)確率達(dá)96.58%,獲得了能對(duì)褶皺紋理圖案精準(zhǔn)識(shí)別的網(wǎng)絡(luò)模型,,實(shí)現(xiàn)具有安全性的防偽目的,。
中圖分類(lèi)號(hào): TP391
文獻(xiàn)標(biāo)識(shí)碼: A
DOI: 10.19358/j.issn.2096-5133.2021.02.007
引用格式: 陳雨,陳桂雄,,周雄圖,,等. 基于DeepLabv3的隨機(jī)褶皺防偽圖案識(shí)別研究[J].信息技術(shù)與網(wǎng)絡(luò)安全,2021,,40(2):39-44.
文獻(xiàn)標(biāo)識(shí)碼: A
DOI: 10.19358/j.issn.2096-5133.2021.02.007
引用格式: 陳雨,陳桂雄,,周雄圖,,等. 基于DeepLabv3的隨機(jī)褶皺防偽圖案識(shí)別研究[J].信息技術(shù)與網(wǎng)絡(luò)安全,2021,,40(2):39-44.
Research on the recognition of anti-counterfeiting pattern based on DeepLabv3
Chen Yu1,,Chen Guixiong1,Zhou Xiongtu1,,2,,Zhang Yongai1,2,,Lin Zhixian1,,2,Wu Chaoxing1,,2,,Guo Tailiang1,2
(1.College of Physics and Information Engineering,,F(xiàn)uzhou University,,F(xiàn)uzhou 350116,China,; 2.Fujian Science & Technology Innovation Laboratory for Optoelectronic Information of China,,F(xiàn)uzhou 350116,China)
Abstract: In view of the problems of anti-counterfeiting technology, such as cloneable, low reliability, and high cost, this paper proposed an identification method for random wrinkle formed by compressive stress caused by the mismatch of thermal expansion index. The paper used DeepLabv3, a edge of deep convolution network classification algorithm, for classification and recognition. Through optimizing the full connectivity layer and setting different neuron nodes, the classification accuracy of recognition network was improved, the training time was reduced, the training accuracy rate was as high as 96.58%, the network model for accurate recognition of wrinkle texture pattern was acquired, and the security purpose of anti-counterfeiting was realized.
Key words : anti-counterfeiting,;deep learning,;DeepLabv3;image classification Artificial Intelligence
0 引言
市場(chǎng)中假冒產(chǎn)品的存在會(huì)對(duì)國(guó)家,、社會(huì)和個(gè)人帶來(lái)巨大經(jīng)濟(jì)損失,,防偽成為應(yīng)用廣泛的反制技術(shù)。由于整個(gè)防偽市場(chǎng)不規(guī)范,,防偽技術(shù)產(chǎn)品水平偏低,,妨礙了市場(chǎng)的健康發(fā)展,公眾對(duì)防偽產(chǎn)品的信任度在降低,。目前,,許多被開(kāi)發(fā)的防偽標(biāo)簽具有物理上不可克隆的特征,如散射表面的隨機(jī)圖案,、隨機(jī)分布的納米顆粒圖案和液晶紋理等,。褶皺圖案是自然界生物體和工程材料領(lǐng)域常見(jiàn)的特殊現(xiàn)象,,是一種微觀的隨機(jī)地形,擁有著廣泛而不可復(fù)制的信息,,在防偽技術(shù)上有廣泛的應(yīng)用前景,。
本文詳細(xì)內(nèi)容請(qǐng)下載:http://forexkbc.com/resource/share/2000003377
作者信息:
陳 雨1,陳桂雄1,,周雄圖1,,2,張永愛(ài)1,,2,,林志賢1,2,,吳朝興1,,2,郭太良1,,2
(1.福州大學(xué) 物理與信息工程學(xué)院,,福建 福州350116; 2.中國(guó)福建光電信息科學(xué)與技術(shù)創(chuàng)新實(shí)驗(yàn)室,,福建 福州350116)
此內(nèi)容為AET網(wǎng)站原創(chuàng),,未經(jīng)授權(quán)禁止轉(zhuǎn)載。