Super-resolution image reconstruction method for geostationary orbit satellite attitude under-sampling measurement
A technology for geostationary orbit satellites and low-resolution images, which is applied in image enhancement, image data processing, graphics and image conversion, etc., and can solve problems such as long acquisition time intervals and inability to use processing methods for super-resolution reconstruction
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
Method used
Image
Examples
Embodiment 1
[0062] This embodiment provides a method for super-resolution image reconstruction under under-sampling measurement of the attitude of a synchronous orbiting satellite, which includes the steps:
[0063] Step 101: Read the time-undersampled low-resolution sequence image I L1 ; Because the frequency of the attitude measurement information obtained by the geostationary orbit satellite is lower than the frequency of the camera image, it is impossible to estimate the two adjacent images according to the imaging formula established by the satellite attitude measurement information while acquiring each low-resolution image The relative displacement between the two, which also seriously affects the effect of the subsequent super-resolution reconstruction algorithm. Therefore, first establish a relative motion model for low-resolution sequence images.
[0064] Step 102: Since the deviation of the sequence image relative to the reference image is caused by the attitude stability of the geos...
Embodiment 2
[0110] Such as figure 2 As shown, the specific implementation method of the present invention is as follows:
[0111] (1) First read the time-undersampled low-resolution sequence image I L1 ;
[0112] (2) The satellite attitude change is a two-dimensional change. Because the camera and the star are considered as a complete rigid body in this patent, only the low-frequency information of its vibration is considered. First, a low-order model that can reflect its movement changes is established according to the following formula , Get the complete image sequence I caused by the basic satellite attitude change L2rough ,
[0113]
[0114] X(t) refers to the displacement of the image in the x direction over time, A refers to the amplitude of the image in the x direction, ω x Refers to the vibration frequency of the image in the x direction, Refers to the phase of the image in the x direction, Y(t) refers to the displacement of the image in the y direction over time, B refers to the ampl...
Embodiment 3
[0153] On the basis of Embodiment 1 and Embodiment 2, this embodiment is an application embodiment.
[0154] Knowing that a certain camera takes images at 0s, 5s, 10s, 25s, and 35s, a low-order model reflecting the vibration changes is established according to the displacement relationship between the captured images as the next state transition matrix, as follows,
[0155]
[0156] According to Kalman two-dimensional filtering, the final image sequence change model is obtained:
[0157] x t +0.5281x t-1 -0.8861x t-2 -0.01886x t-3 +0.8568x t-4 +0.2776x t-5
[0158] =ε t +0.299ε t-1 -1.117ε t-2 +0.2802ε t-3 +0.9399ε t-4
[0159] y t +0.2337y t-1 -1.114y t-2 -0.9148y t-3 +0.01707y t-4 +0.7262y t-5 +0.1876y t-6
[0160] =ε t +0.2459ε t-1 -1.322ε t-2 -1.096ε t-3 -0.5922ε t-4 +0.9435ε t-5 +0.5864ε t-6
[0161] +0.4674ε t-7 -0.1208ε t-8 -0.6214ε t-9
[0162] According to the above two formulas, the image data of 0s, 5s, 10s, 15s, 20s, 25s, 30s, 35s are obtained, and then these images are registe...
PUM
Login to View More Abstract
Description
Claims
Application Information
Login to View More 


