分数布朗运动模型的人造目标检测

Artificial Target Detection based on Fractional Brownian Movement Model

对光电目标进行自动检测与识别是导弹末端智能制导的一项关键技术,分形理论的出现为解决这个问题提供了新的技术途径。分形维计算的稳定性问题制约着其在实际工程中的应用。基于图像分形维计算的分数布朗运动模型法,提出了一种能够从自然背景中检测出人造目标的方法。实验证明,对于光电目标检测,给出方法的检测结果稳定,是一种非常有前途的方法。

Automatic detection and recognition of electro-optical target is a key technology in the intelligent missile terminal guidance. Fractal theory provides a new technical way. In the application of practical engineering, stability of the fractal dimension computation still limits its application of the theory. Based on the fractional Brownian movement model, a practical approach is presented to detect an artificial target from natural background. Experiments show that the method proposed in this paper is a promising detection method of the electro-optical target.