Image forensics has become increasingly important in our daily lives. As a fundamental type of forgeries, Copy-Move Forgery Detection (CMFD) has received significant attention in the academic community. Keypoint-based algorithms, particularly those based on SIFT, have achieved good results in CMFD. However, the most of keypoint detection algorithms often fail to generate sufficient matches when tampered patches are present in smooth areas. To tackle this problem, we introduce entropy images to determine the coordinates and scales of keypoints, resulting significantly increasing the number of keypoints. Furthermore, we develop an entropy level clustering algorithm to avoid increased matching complexity caused by non-ideal distribution of grayscale values in keypoints. Experimental results demonstrate that our algorithm achieves a good balance between performance and time efficiency.

翻译：图像取证在我们的日常生活中日益重要。作为伪造的基本类型之一，复制-移动伪造检测（CMFD）在学术界受到了广泛关注。基于关键点的算法，特别是基于SIFT的算法，在CMFD中取得了良好效果。然而，当篡改区域位于平滑区域时，大多数关键点检测算法往往无法生成足够的匹配点。为解决此问题，我们引入熵图像来确定关键点的坐标和尺度，从而显著增加关键点数量。此外，我们开发了熵水平聚类算法，以避免由于关键点灰度值分布不理想而导致的匹配复杂度增加。实验结果表明，我们的算法在性能和时间效率之间实现了良好平衡。