Research On Ground Targets Recognition Positioning Method For Forward-Looking Sequence Of Images

In different time and weather condition, properties of images form infrare sensor aresignificant different. When target recognizition is influenced by complex factors, e.g thetarget is covered by cloud or out of the field of view,light effect causes the image contrastlow, existence of false target or repeat mode, ground scene changing, influence fromcomplex factors such as noise and blur and so on,reduce the recognition rate and robust.Inthe resource-constrained conditions,how to improve the rate and real-time of ATR systemis a major problem in modern navigation technology area.This paper presents an approach to the issues of target recognizition under the compleximaging environmental and proposes a kind of ATR strategy and a method forforward-looking sequence of images, as target recursive forecast recognition strategy forsequence of images based on flight attitude parameters and target recognition method forsequence of images based on context of the landmarks.As the high recognitionrate,real-time and wide using range, they can be used to effectively improve the stabilityand positioning accuracy of target recognition. So it provides an efficient and reliablerecognition strategy for precise navigation.It also studies on the matching of multisensor images,and proposes a matching methodof image structure detection based on implicit similarity measure.It combines the grayinformation with the structure feature information, and it is a good solution to thematching of multisensor images.It can be used in the image matching betweeninfrare,SAR and optic images.It is used for the initial acquisition in the target recognitionmethod in orde to improve the detection accuracy.Last the algorithm implementation,design and optimization in DSP is also introduced.A large number of simulation experiments are proposed to analyze the adaptability andtrack accuracy of the algorithm in a variety of complex imaging environment,such astarget is covered by cloud,in different time condition, ground scene changing and so on.