The work aims to design an image copy-move forgery detection algorithm based on Harris operator coupling intensity to solve the problem that most existing copy-and-paste tamper detection algorithms mainly rely on measuring the structural characteristics of the image to achieve tamper detection, thus ignoring the image strength characteristics, making the algorithm can not adapt to rotation and other tamper detection methods, resulting in the detection results of leakage detection and error detection. The Harris operator was used to extract the feature points accurately. A circular feature region was constructed by feature points, and the Zernike moments of the region were obtained. The feature points were described by the size of the Zernike moments. Zernike moments of different orders were used to describe the intensity and texture features of the image. The matching model was constructed by the size of Zernike moments of different orders to roughly match the image features. The RANSAC method was introduced to further realize the precise matching of image features. The feature area was connected by morphological corrosion and expansion operation to determine the tampering area. From the experimental results, the proposed algorithm had high detection accuracy and roughness and still had better detection results under the noise and rotation, compared with the existing image forgery detection scheme. The proposed technique possesses high forgery detection accuracy and has certain reference value in the fields of image watermarking and information security.