Application Research On Information Fusion Technology Of Radar Sensor And Infrared Sensor In UAV

With growing complexity of the battlefield environment today, as the UAV technology improving, while the reconnaissance capability, monitoring capability and strike capability of the UAV are dependent on the development of the sensor technology. But relative to the current battlefield,The information of the single sensor cannot satisfy the combat requirement of high standard on the modern battlefield. So this subject raises the multi-sensor information fusion technology. Currently the methods of the information fusion are many, because of the effective performance of the D-S evidence theory for the information fusion, so this article selects the D-S evidence theory as the method of the information fusion. But the original D-S evidence theory can not handle the conflict evidence effectively, with the certain circumstances, fusion is failure, the original D-S evidence theory is improved very well aiming at this problems, make it have effective integration ability under any circumstances. In addition,it can enhance fusion result on the basis of the improving of the previous to this problem, made the improved D-S evidence theory had a wider range of applications and higher fusion result.In this paper, It regard the uncertainty reasoning of artificial intelligence as the method of information fusion, and improved the original D-S evidence theory to effectively handle conflict evidence, in order to have the better result of the information fusion.The paper have the two aspects in the image processing research: It have the operation on regional feature extraction with the Hu seven constant distance for the SAR image, and then etermined the recognition framework and calculated basic probability distribution; It have the operation on the edge feature extraction with the improved Sobel operator for the infrared sensor, and then etermined the recognition framework and calculated basic probability distribution. Then this paper had a wholly overview about the information fusion operating environment. Finally it is the simulation experiment, it proved that the methods and the improved algorithm of this article have the strong feasibility and availability by experiments.