中文摘要：

针对局部窗口K分布检测算法运算速度慢、计算效率低的问题,提出了一种基于局部窗口K分布的快速舰船目标检测算法。该算法首先采用迭代分割算法对原始合成孔径雷达（SAR）图像进行预筛选处理,根据预筛选选出潜在目标,在原始SAR图像中剔除潜在目标像素;然后利用背景图像计算二阶和四阶积分图像,在每一个像素点处采用滑动窗口的方式,在积分图像中进行加减计算确定所在位置的二四阶矩并估计K分布的参数;其次,确定概率密度函数后,通过求解函数得到检测阈值,根据检测阈值确定感兴趣区域;最后,通过模糊差影的鉴别方法对目标中的虚警目标进行进一步剔除,进而完成检测。通过实测SAR图像检测实验,积分算法与局部窗口的K分布算法相比将运算所需时间降低了50%,基于模糊差影的鉴别算法将品质因素由44.4%提高到100%。所提算法既保证了算法的实时性,又提高了检测的精度,在进行SAR舰船自动检测方面具有一定的应用价值。

英文摘要：

Aiming at the problem of low efficiency and low computational efficiency of local window K-distribution detection algorithm,a fast ship target detection algorithm based on local window K-distribution was proposed. Firstly,the original Synthetic Aperture Radar（ SAR） image was selected by the iterative segmentation algorithm,and the potential target pixels in the original SAR image were removed according to the pre-selection. Then two-order and four-order integral images were calculated by using the sliding window at each pixel in the background images. Two-order and four-order moments of the K-distribution were calculated in the integral image in order to estimate the parameters of K-distribution. Secondly,the detection threshold was determined by solving the probability density function and the regions of interest were obtained according to the threshold. Finally,the false alarm target was detected by the method of fuzzy difference. The detection experiment using the real SAR image show that the running time of the algorithm is reduced by 50% compared with the local window K-distribution algorithm,and the quality factor is improved from 44. 4% to 100%. The proposed algorithm not only ensures the real-time performance of the algorithm,but also improves the detection accuracy,and it has a certain application value in the automatic detection of SAR ship.