红外烟幕遮蔽条件下军用目标热像图的增强与判别

【作者】 岳贤军；

【导师】 陈文建； 迟泽英；

【作者基本信息】 南京理工大学 ， 物理电子学， 2003， 硕士

【摘要】 为了解决红外烟幕遮蔽条件下军用目标热像的增强与判别以及进一步减少烟幕对红外目标的遮蔽时间问题，本文根据烟幕遮蔽下目标热像的特征，对目标热像利用基于二维直方图的灰度变换方法进行了灰度变换，增强了目标热像与烟幕背景的对比度。根据烟幕所具有的特征以及分形理论适合于描述自然物体和具有不规则形状的物体的特性，把分形理论引入到本课题中，用分形理论来描述烟幕和目标热像图，寻找到分形维数作为识别烟幕背景中目标的分形特征。 本文采用经典的分形布朗运动模型(FBM)来描述烟幕和目标，由于烟幕可视为一自然物体，而目标可视为规则的人造物体，二者的分形维数存在较大的差别。所以本文选择了分形维数作为区分烟幕背景和目标的分形特征。 本课题中采用Visual C++平台，建立起了相应的软件，结合数字图像处理技术，利用基于分形理论的目标识别算法对“9711烟幕外场”试验数据进行了处理，结果表明该方法能够显著地减少烟幕对目标的遮蔽时间，并能在各种烟幕条件下正确的检测出目标。更多还原

【Abstract】 To recognize the infrared targets in infrared smokescreen and decrease shade-time of smokescreen, in this paper, the infrared target image in smokescreen was processed with the method of gray level transformation basing on two-dimension histogram according to the characteristics of infrared target image in infrared smokescreen. Basing on the characteristic of infrared smokescreen and of fractal theory that adapts to describing natural objects and the objects that have irregular shape, in this paper, fractal theory was applied to describing the infrared smokescreen and infrared target to be recognized. And the fractal characteristic of fractal dimension is found to recognize the infrared target in infrared smokescreen.In this paper, the classical model- Fractal Brown movement Model was applied to describing the smoke and target. In that smoke can be regard as natural object, and the target to be recognized is regular manmade object, the difference of fractal dimension between them is obvious. We applied fractal dimension as a characteristic to recognize the target.In the end, Visual C++ platform was use to build the corresponding software. We applied the software and digital image processing technology to process the image data of 9711-outdoor-experimentation. The result showed that the method could decrease the shade-time of smokescreen and could find out the target in smokescreen under any smoke conditions.更多还原