基于CNN的计算机生成图像识别方法
|
摘要
针对计算机生成图像(Computer Generated images, CG)与真实照片(Photograpgh, PG)识别率不高的问题,该文提出了一种改进的卷积神经网络方法来实现CG与PG的识别.该方法首先对识别问题进行卷积神经网络二分类建模,并选择VGG-19网络结构作为基础,建立不同的模型.该方法创新性地引入迁移学习,节省训练时间和大量计算资源,最后使用softmax分类器进行分类.实验结果表明,该文方法对PG图像的识别准确率达到92%.与其他方法比较,该文方法识别准确率最高,说明该文方法具有可行性与有效性.
In order to solve the problem low recognition rate for Computer Generated images(CG) and Photographs(PG), an improved convolution neural network method is proposed to realize the recognition of CG and PG. This method first set up the two-classification model of the convolution neural network for the recognition problem and the VGG-19 network structure is selected as the basis to establish different models. This method innovatively introduces migration learning and saves training time and massive computing resources. Finally, softmax classifier is used to classify. The experimental results show that the accuracy of the proposed method for PG image recognition is up to 92%, and the recognition speed is faster. Compared with other methods, the method has the highest recognition accuracy and demonstrates the feasibility and effectiveness of the proposed method.
In order to solve the problem low recognition rate for Computer Generated images(CG) and Photographs(PG), an improved convolution neural network method is proposed to realize the recognition of CG and PG. This method first set up the two-classification model of the convolution neural network for the recognition problem and the VGG-19 network structure is selected as the basis to establish different models. This method innovatively introduces migration learning and saves training time and massive computing resources. Finally, softmax classifier is used to classify. The experimental results show that the accuracy of the proposed method for PG image recognition is up to 92%, and the recognition speed is faster. Compared with other methods, the method has the highest recognition accuracy and demonstrates the feasibility and effectiveness of the proposed method.
引文
[1] 申铉京,李梦臻,吕颖达,等.基于LBC的计算机生成图像盲鉴别算法 [J].计算机科学,2015,42(6):135-138,161.
[2] 敖道敢.无监督特征学习结合神经网络应用于图像识别 [D].广州:华南理工大学,2014.
[3] DE REZENDE E R S,RUPPERT G C S,CARVALHO T.Detecting Computer Generated Images with Deep Convolutional Neural Networks [C]//2017 30th SIBGRAPI Conference on Graphics,Patterns and Images (SIBGRAPI).Niteroi:IEEE,2017.
[4] 刘内美.高识别率的鲁棒ART快速图像识别算法 [J].西南师范大学学报(自然科学版),2016,41(8):106-113.
[5] PENG F,ZHOU D L,LONG M,et al.Discrimination of Natural Images and Computer Generated Graphics Based on Multi-Fractal and Regression Analysis [J].AEU-International Journal of Electronics and Communications,2017,71:72-81.
[6] TOKUDA E,PEDRINI H,ROCHA A.Computer Generated Images vs.Digital Photographs:A Synergetic Feature and Classifier Combination Approach [J].Journal of Visual Communication and Image Representation,2013,24(8):1276-1292.
[7] CONOTTER V,BODNARI E,BOATO G,et al.Physiologically-Based Detection of Computer Generated Faces in Video [C]//2014 IEEE International Conference on Image Processing (ICIP).Paris:IEEE,2014.
[8] 张震,杨宇豪.基于Benford模型的自然图像与计算机生成图像的鉴别 [J].北京工业大学学报,2013,39(6):930-935.
[9] TAN D Q,SHEN X J,QIN J,et al.Detecting Computer Generated Images Based on Local Ternary Count [J].Pattern Recognition and Image Analysis,2016,26(4):720-725.
[10] 魏晓聪,林鸿飞.面向迁移学习的文本特征对齐算法 [J].计算机工程,2017,43(2):215-219,226.
[11] SHIN H C,ROTH H R,GAO M C,et al.Deep Convolutional Neural Networks for Computer-Aided Detection:CNN Architectures,Dataset Characteristics and Transfer Learning [J].IEEE Transactions on Medical Imaging,2016,35(5):1285-1298.
[2] 敖道敢.无监督特征学习结合神经网络应用于图像识别 [D].广州:华南理工大学,2014.
[3] DE REZENDE E R S,RUPPERT G C S,CARVALHO T.Detecting Computer Generated Images with Deep Convolutional Neural Networks [C]//2017 30th SIBGRAPI Conference on Graphics,Patterns and Images (SIBGRAPI).Niteroi:IEEE,2017.
[4] 刘内美.高识别率的鲁棒ART快速图像识别算法 [J].西南师范大学学报(自然科学版),2016,41(8):106-113.
[5] PENG F,ZHOU D L,LONG M,et al.Discrimination of Natural Images and Computer Generated Graphics Based on Multi-Fractal and Regression Analysis [J].AEU-International Journal of Electronics and Communications,2017,71:72-81.
[6] TOKUDA E,PEDRINI H,ROCHA A.Computer Generated Images vs.Digital Photographs:A Synergetic Feature and Classifier Combination Approach [J].Journal of Visual Communication and Image Representation,2013,24(8):1276-1292.
[7] CONOTTER V,BODNARI E,BOATO G,et al.Physiologically-Based Detection of Computer Generated Faces in Video [C]//2014 IEEE International Conference on Image Processing (ICIP).Paris:IEEE,2014.
[8] 张震,杨宇豪.基于Benford模型的自然图像与计算机生成图像的鉴别 [J].北京工业大学学报,2013,39(6):930-935.
[9] TAN D Q,SHEN X J,QIN J,et al.Detecting Computer Generated Images Based on Local Ternary Count [J].Pattern Recognition and Image Analysis,2016,26(4):720-725.
[10] 魏晓聪,林鸿飞.面向迁移学习的文本特征对齐算法 [J].计算机工程,2017,43(2):215-219,226.
[11] SHIN H C,ROTH H R,GAO M C,et al.Deep Convolutional Neural Networks for Computer-Aided Detection:CNN Architectures,Dataset Characteristics and Transfer Learning [J].IEEE Transactions on Medical Imaging,2016,35(5):1285-1298.