A novel method is proposed to identify digital image origin with the correlation theory. Digital images form digital cameras contain relatively stable residual noise due to the imperfection of imaging of imaging sensors. After deriving residual noise from an image through wavelet filtering,a reference pattern of noise for a digital camera is extracted by averaging the residual noise from multiple images. By comparing the correlation between the residual noise of the image to be evaluated and the reference pattern with a predetermined threshold,the origin of the image can be determined through. Bayesian decision theory with minimum-error-rate is used to derive the threshold. Experiment results show the effectiveness of the proposed method.