基于DWT和形态学滤波的图像伪造检测方法

Image Forgery Detection Based on DWT and Morphological Filtering

摘要:

针对图像中复制-移动和拼接形式的图像伪造检测,提出一种基于离散小波变换(DWT)和形态学滤波的图像伪造检测方法。首先,将图像转换为灰度图,通过应用DWT获得LH、HL和HH子带。然后,通过阈值判断来获得伪造图像区域的边缘,并通过形态学滤波来连接边缘使其清晰化。最后,提取伪造区域的SIFT特征,并通过相似性检测来寻找图像中与伪造区域相似的区域,以此来确定伪造类型。实验结果表明,该方法能够准确检测出伪造区域和伪造类型。

Abstract:

For the issues that the image forgery detection in the form of copy-move and splicing, an image forgery detection method based on discrete wavelet transform (DWT) and morphological filtering is proposed. Firstly, the image is converted to grayscale and the LH, HL, and HH subbands are obtained by applying a DWT. Then, the edge of the forgery image area is obtained by the threshold judgment, and the edges are sharpened through morphological filtering. Finally, the SIFT feature of the forged region is extracted and the similar region in the image is searched through the similarity detection. Experimental results show that this method can accurately detect forgery regions and forgery types.