Electronic Image Stabilization Technology Of Airborne Infrared Imagery

Due to the country's vigorous development of the marine economy,the increasing number of maritime activities and the increasingly complex traffic environment,coupled with the impact of anomalous climate changes,have resulted in continual sea accidents.In order to improve the existing marine search and rescue capability in our country,it is necessary to develop a stable,fast and effective rapid search system for airborne sea surface distress targets.Infrared detection with penetrating ability,detection range,can be monitored around the clock and other advantages,in the night and severe climatic conditions clear imaging,to achieve the target detection and monitoring.Airborne infrared search systems installed on rescue helicopters can greatly enhance the search capability of distress targets on the surface of the sea.However,due to the airborne environment,the instability of the aircraft can cause the video image to jitter,blur and affect the target identification,location and tracking.The use of electronic image stabilization method can effectively solve the above problems.This paper mainly studies the electron image stabilization technology of airborne infrared images.On the basis of analyzing the characteristics of airborne infrared images,global motion estimation,motion filtering and motion compensation are carried out to complete the jittering infrared video frame processing.Motion estimation is based on the detection of global motion vectors between two adjacent video frames of a dithered video and is recovered by motion compensation.The accuracy of motion estimation directly affects the quality of the video.The speed of motion estimation directly affects the processing time of the imaging system.Therefore,this paper improves the optical flow tracking algorithm,the use of the infrared image feature points for optical flow tracking,reducing the amount of computation.For the problem of large inter-frame movement and poor optical tracking performance,a pyramid optical flow tracking algorithm based on multi-scale is proposed in this paper.By using the constraint of the interlayer optical flow vector,the accuracy of the optical flow is improved.The motion filtering can preserve the active component of jitter between frames.The Kalman filter is used in this paper.The experimental results show that the motion filtering algorithm can obtain smooth motion parameters.Motion compensation uses a stable image model for compensation transformation.After the image sequence is processed,the information of some areas is lost.In this paper,a method based on neighborhood compensation is used to compensate for the black edges.In this paper,the validity of the algorithm is analyzed experimentally.Experiments show that this algorithm can process jitter video sequences in real time.