Motion Imgae Restoration Based On OPTO-Electronic Hybrid Joint Transform Correlator

With the rapid development of airborne remote sensing technique, high resolution image is required more than ever. These high resolution images can satisfy the requirement of digital urban planning, military reconnaissance, natural disaster monitoring and early warning, geographical information system. However, there are many reasons which caused the captured image to be blurred, such as movement, environment noise, transfer degradation. In this paper, motion blur is mainly discussed, which is caused by image motion between target and camera.First of all, the investigation background and its significance are introduced. And then the principle of motion image restoration is showed, such as degradation model, four classical restoration methods and quality assessment of restoration. The restoration of motion blurred image is simulated, the result shows that the more accurate point spread function (PSF), the better restoration image quality will be gotten.Next, according to current study, an effective restoration system is proposed based on opto-electronic hybrid correlator. In this system, one primary camera captures a high resolution but blurred image during a long exposure time, while the high speed camera with low resolution captures image sequences at the same time. The movement information is included in this image sequence. And then these sequences are loaded into JTC system sequentially. The movement vectors can be obtained by optical correlator calculation, then a modified centroid method is used to locate the correlation peaks. The accurate PSF can be calculated based on the movement vectors. In the end, the constrained least square algorithm is used to restore blurred image. However, these sequences always have the problem of low contrast ratio, complicated background, which will make the correlative peak undetectable. Three modified methods are introduced to improve the correlative output. Simulation and optical experiment results show that Laplace transform can improve correlative peak greatly.At last, the motion image restoration based on JTC is simulated and its optical experiment is set up. Experiment result shows that this method can calculate the accurate PSF, and it can restore the motion blurred image effectively.