Joint Detection System Of Key Technologies

With the war of optical countermeasure becoming more and more fierce recently, base on reconnaissance and anti-reconnaissance, electrical jamming, reconnaissance and surveillance, precision attack become the focus matter of a war in modern times. In a battle, spatial war is the most fierce part, and it often plays a fateful role.For target detection, the information quantity acquired by single detection way or single detector can't meet the demand of war. Along with the detection techniques and data fusion techniques improving and developing gradually, now joint detection using multi-wavelength and multi-way becomes the primary way in battle field.To solve the problems, for primary spatial targets, this paper brings forward joint detective system project, so we can meet the all-day, all-weather and high precision detection demand in battle field condition. The primary content of this paper are:The paper makes analyses with spectrum characteristics of spatial targets (fixed-wing aircrafts, military helicopters and precision guided bombs) and background, and merits and demerits of active and passive detection, brings up a joint detective system with visible band and NIR band for passive detection and 1.06μm lidar for active diction, makes a conception for optic joint detective system.The paper makes emphasize on target detection and tracking algorithm, brings up a system base on multi-sensor fusion for target detection and tracking system, uses adaptive threshold frame difference algorithm for beforehand detection, uses adaptive window algorithm to track with the target, and uses wavelet algorithm for multi-sensor fusion, makes image registration before image fusion.The paper makes simulation experiment for adaptive threshold frame difference algorithm, and makes comparison with general threshold frame difference algorithm, makes a lot of simulation experiment for image fusion using different wavelet filters, different decomposition level. The paper makes performance evaluations of fusion images using entropy, cross entropy, deviation and joint PSNR, gets some useful conclusions through experiment data.