Satellite image jitter restoration using magnetohydrodynamics angular rate sensor and total variation regularization

Satellite jitter seriously affects the acquisition, pointing, and tracking performance of the remote sensing imaging system. Reliable detection and image restoration methods are required to minimize the unfavorable jitter. In this work, a magnetohydrodynamics angular rate sensor is used to measure the platform jitter within a wide frequency range of 2–1000 Hz. In accordance with the measured data, the point spread function is constructed. A non-blind image restoration algorithm based on the improved Total Variation (TV) regularization method is proposed. The blurred image is first divided into texture region and smooth region based on the local entropy. Binarization operations are then performed on the texture region and the smooth region. The smooth domain and the deblurred image restored by the traditional TV regularization algorithm are multiplied by pixels. The acquired image determines the smooth information of the restored image. The texture region is multiplied with the one recovered by the proposed improved TV regularization method to guarantee the image details. The blurred image is processed using the Canny operator to obtain the image gradient weights for TV regularization rather than gradient minimization. Finally, the acquired images are superimposed to obtain the ultimate restored image. Extensive experiments show that the proposed method performs better than the traditional TV regularization algorithm.