Skip to Main Content
Due to different levels and characteristic deformation existing in any images, it is difficult to select an appropriate model to correct effectively. Considering BP neural network is a high and nonlinear complicated system which can approach at any precision, we attempt to use it for image geometric correction. In this paper, errors correction of scanning map by BP neural network as an example was discussed. The result shows that the accuracy of scanning map corrected by the BP neural network is higher than by normal function fitting.