| Laser radar and passive infrared imaging can get the image of target, the performance of imaging system and the final image sequence are effected by the carrier's attitude and fluctuation, so the video image sequence of target is unstable and blur. As the comprehensive application of video camera system, the require of stabilization of video is more and more important. The effect of image stabilization has an immediate effect on the target identification. Electronic image stabilization helps to process the video image sequence, and hold the each target more in the image on the whole.This paper gives the basic method of the electronic image stabilization, electronic image stabilization system usually has two parts: motion estimation and motion compensation. During the part of motion estimation, with the help of image processing and information extract, we get the local motion estimation and global motion estimation, then make sure to find the motion vectors between the frames. So the motion compensation stabilize the original video image sequence. Based on this method, processing includes representative point matching, projection algorithm, block matching, bit plane matching. After the processing of video image sequence, we compare the effect of image stabilization and the superiority.After selecting a picture, we emulate the original image sequence for the experiment. And the real image sequence is got by the camera. This paper use the program of electronic image stabilization compiled by myself to process the emulated and real video image sequence, and realize the processing of the four image stabilization algorithm. Finally I use the quality factor PSNR to evaluate the accuracy of the algorithm, evaluate the stabile degree of the processed video image sequence and their application extent.The result proves that the operator in this article can remove flutter and improve the stability of video image sequence, performance stabile and accurately. |
PDF Full Text Request