论文部分内容阅读
主要研究高分辨率SAR图像中,基于恒虚警率(Constant False Alarm Ratio,CFAR)的输电铁塔检测.利用在高分辨率SAR图像中,球不变随机向量(Spherically Invariant Random Vector,SIRV)模型能够更好地拟合森林和建筑等复杂区域的特点,通过SIRV模型增强来提高CFAR检测的准确率.同时,针对逐点的恒虚警率检测和SIRV模型参数估计速度慢的情况,提出了分级检测的方法,提高了检测速度.实验结果表明,改进后的算法能够实现对输电铁塔快速而准确的检测.
This paper mainly studies the transmission tower detection based on Constant False Alarm Ratio (CFAR) in high-resolution SAR images.Using the Spherically Invariant Random Vector (SIRV) model in high-resolution SAR images, Which can better fit the characteristics of complex regions such as forests and buildings and improve the accuracy of CFAR detection by SIRV model enhancement.At the same time, for point-by-point constant false alarm rate detection and SIRV model parameter estimation speed is slow, Grading detection method to improve the detection speed.The experimental results show that the improved algorithm can achieve fast and accurate detection of transmission tower.