基于改进深度神经网络的数字图像篡改检测方法  

Digital image tampering detection method based on improved deep neural network

作　　者：韩建波[1] 王伟[2] HAN Jianbo;WANG Wei(Hebei School of Science and Technology,Baoding 071000,Hebei China;Tianjin University of Science and Technology,Tianjin 300457,China)

机构地区：[1]河北省科技工程学校,河北保定071000 [2]天津科技大学,天津300457

出　　处：《粘接》2024年第12期139-141,149,共4页Adhesion

基　　金：河北省职业教育科学研究“十四五”规划2022年度项目(项目编号:JZY22110)。

摘　　要：针对现有数字图像内容篡改检测方法存在的检测精度低、检测类型单一等问题,基于现有数字图像篡改技术,提出一种用于数字图像篡改检测的改进深度神经网络。利用Scharr算子提取图像的边缘信息,通过灰度共生矩阵将不同的边缘矩阵统一为相同的大小,在输入改进的深度神经网络中进行多篡改检测。通过试验与常规方法进行了比较分析,验证了所提方法的优越性。结果表明,所提出方法与常规图像篡改检测方法相比,在数字图像篡改检测中具有更优的检测准确率和检测速度,检测准确率达到98.25%,检测速度达到12.99 FPS。In response to the problems of low detection accuracy and single detection type in existing digital image content tampering detection methods,an improved deep neural network for digital image tampering detection was proposed based on the existing digital image tampering technology.The Scharr operator was used to extract the edge information of the image,and the different edge matrices were unified to the same size through the grayscale symbiosis matrix,and the multi-tamper detection was carried out in the input improved deep neural network.The superiority of the proposed method was verified by comparing the experimental method with the conventional method.The results indicated that,compared with conventional image tamper detection methods,the proposed method had better detection accuracy and speed in digital image tamper detection,with a detection accuracy of 98.25%and a detection speed of 12.99 FPS.

关 键 词：数字图像 篡改技术 深度神经网络 Scharr算子 灰度共生矩阵 

分 类 号：TP274[自动化与计算机技术—检测技术与自动化装置]