The work aims to put forward an image forgery detection algorithm based on gradient histogram coupling density measurement model, for the purpose of solving the problem of poor detection accuracy and robustness induced by the current image forgery detection algorithm that easily loses the color information in the process of identifying the contents. Firstly, the RGB color image mapping model was introduced to obtain the color invariants of the image. With the image color invariants as the inputs, the Hessian algorithm was used to detect the feature points of the image. Then, the four-level window was constructed with feature points as the center, and the low dimensional feature descriptor was formed by finding the gradient accumulation value in the window; and the similarity measurement model was constructed with the gradient histogram corresponding to the feature points to match the feature points. Finally, the Euclidean distance was used to construct the density measurement model, which was used to classify the feature points, and then the forgery detection was completed. The simulation results showed that, compared with the current image forgery detection algorithm, the proposed algorithm had higher detection accuracy which was up to 99.6%. The proposed algorithm has higher forgery detection accuracy and robustness, and it has good application value in image information, packaging, printing and other fields.