摘要
由于一幅图像中的信息非常复杂,所以对图像的处理需要考虑图像的许多特性。利用这些特性,伪造者可以通过构建一个或者多个组合来操作一张数码照片,从而使它们的检测更加困难。为了对伪造图像进行识别检测,在此提出了一种新的图像检测算法。该算法将图像分割成若干区域,通过对比比对区域的相似度来确认图像中是否存在复制区域,从而判断该图像是否伪造。在测试部分,通过案例测试证明了提出的方法具有较好的鲁棒性。
Because the information in an image is very complex,the processing of the image needs to consider many characteristics of the image.Using these features,counterfeiters can manipulate a digital photo by constructing one or more combinations,making their detection more difficult.In order to identify and detect forged images,a new image detection algorithm is proposed here.This algorithm divides the image into several regions and compares the similarity of the regions to confirm whether there is a duplicated region in the image,thereby judging whether the image is forged.In the test part,case tests prove that the proposed method has good robustness.
作者
张小娜
Zhang Xiaona(Sichuan Tianyi College,Mianzhu 621000,China)
出处
《单片机与嵌入式系统应用》
2021年第10期49-53,共5页
Microcontrollers & Embedded Systems
基金
国家自然科学基金(61802320)。
