Copy-move is one of the most common image forgeries, wherein one or more region are copied and pasted within the same image. The motivations of such forgery include hiding an element in the image or emphasizing a particular object. Copy-move image forgery is more challenging to detect than other types, such as splicing and retouching. Keypoint based copy-move forgery detection extracts image keypoints and uses local visual features to identify duplicated regions, which exhibits remarkable performance with respect to memory requirement and robustness against various attacks. However, these approaches fail to handle the cases when copy-move forgeries only involve small or smooth regions, where the number of keypoints is very limited. Also, they generally have higher time costs owing to complex feature descriptor and more error matching points. To tackle these challenges, we propose a fast and effective copy-move forgery detection method through adaptive keypoint extraction and processing, introducing fast robust invariant feature, and filtering out the wrong pairs. Firstly, the uniform distribution keypoints are extracted adaptively from the forged image by employing the fast approximated LoG filter and performing the uniformity processing. Then, the image keypoints are described using fast robust invariant feature and matched through the Rg2NN algorithm. Finally, the falsely matched pairs are removed by employing the segmentation based candidate clustering, and the duplicated regions are localized using optimized mean-residual normalized production correlation. We conduct extensive experiments to evaluate the performance of the proposed scheme, in which encouraging results validate the effectiveness of the proposed technique, in comparison with the state-of-the-art approaches recently proposed in the literature.

中文翻译：

---

一种快速、高精度的图像复制移动伪造检测方法

复制移动是最常见的图像伪造之一，其中一个或多个区域被复制并粘贴到同一图像中。这种伪造的动机包括隐藏图像中的元素或强调特定对象。复制移动图像伪造比其他类型（例如拼接和修饰）更难检测。基于关键点的复制移动伪造检测提取图像关键点并使用局部视觉特征来识别重复区域，在内存需求和对各种攻击的鲁棒性方面表现出卓越的性能。然而，这些方法无法处理复制移动伪造仅涉及小区域或平滑区域的情况，其中关键点的数量非常有限。此外，由于复杂的特征描述符和更多的错误匹配点，它们通常具有更高的时间成本。为了应对这些挑战，我们提出了一种快速有效的复制移动伪造检测方法，通过自适应关键点提取和处理，引入快速鲁棒不变特征，并滤除错误对。首先，通过采用快速近似LoG滤波器并进行均匀性处理，从伪造图像中自适应地提取均匀分布的关键点。然后，使用快速鲁棒不变特征描述图像关键点，并通过 Rg2NN 算法进行匹配。最后，通过采用基于分割的候选聚类去除错误匹配的对，并使用优化的均值-残差归一化生产相关性来定位重复区域。我们进行了广泛的实验来评估所提出方案的性能，