Abstract: Image forgery has been a serious issue in recent years.Traditional authentification methods depend on experience excessively,and lack strict and consistent standards.In this paper,a splicing image authentification method based on color temperature estimation is proposed.Our method uses an improved chromatic aberration calculation algorithm to estimate color temperature of key areas in images.Then distances between these regions are computed with the results of color temperatures.After the above procedure,the distance is compared with a threshold value based on experiments to find whether color temperatures are inconsistent among these images.Experiments demonstrated that our method can verify splicing images exactly,and provide scientific evidences for image authentification.