Aerial Image Displacement Measurement And Simulation Based On Improved Joint Transform Correlation Algorithm

During aerial image acquisition, any tiny disturbance will make the image motion of the target image on the focal plane of the camera, thus limiting the image resolution and reducing the image quality. In order to improve the image quality, it is necessary to compensate the image motion, but the image motion should be accurately measured. The correlation operation of two time series images is performed by the JTC(joint transform correlation) algorithm, the displacement of the correlation peak is measured, and the measurement of image motion is realized. However, traditional JTC algorithm aiming at the large scale displacement of the pixel level, the principle and experimental study of sub-pixel displacement is not sufficient. In addition, there is a big error between the displacement measurement accuracy of pixel level and the actual displacement. Therefore, this paper makes an improvement on the JTC algorithm based on the high resolution aerial image, the improved algorithm can measure the sub-pixel displacement of high resolution aerial image with high accuracy. The main work is as follows:(1) The traditional edge detection algorithm can get effective edge detection results if the edge feature of the image is clear enough. For aerial image, the edge profile is not obvious and the noise is complex, the traditional edge detection algorithm is difficult to accurately detect the edge feature. So, this paper uses the fuzzy edge detection algorithm to replace the traditional edge detection algorithm.(2) There is a strong autocorrelation peak in the relevant output results, and its presence will affect the accurate measurement of the cross correlation peak of the displacement information. So, the improved method is used to enhance the joint power spectrum, which is used to enhance the useful information of the power spectrum and get the sharp cross correlation peak.(3) The MATLAB simulation results show that the improved algorithm not only can measure the displacement with accuracy of sub-pixel level, but also has obvious anti-noise ability.