论文部分内容阅读
针对海水背景对舰船目标检测的干扰问题,提出了1种基于多通道Gabor滤波模糊综合评价融合方法来抑制海水背景,增强舰船目标区域,并实现舰船目标的检测和提取。首先对图像进行多通道Gabor滤波,得到多幅滤波增强输出图像;其次,定义了3种滤波图像增强效果评价指标,并为输出图像建立模糊评价矩阵;再次,根据模糊评价矩阵计算出各输出图像的模糊综合评价值,并选出各通道滤波增强效果最优的输出图像,作为该通道的滤波输出显著图像;最后,通过各显著图像的模糊评价值,计算对应的融合权重,并对这些输出显著图像进行加权叠加融合,得到舰船目标融合增强图像并进行检测。实验结果表明,本研究提出的方法能够自适应选取具有较好背景抑制效果和舰船目标区域增强效果的Gabor滤波输出图像进行融合,融合后的图像能够有效增强舰船目标的显著性。与现有的基于多通道Gabor滤波的舰船目标检测方法相比较,本研究提出的舰船目标检测算法能够有效减少目标检测的虚警率,提高检测的正确率。
Aiming at the interference of seawater background to the detection of ship targets, a fuzzy comprehensive evaluation fusion method based on multi-channel Gabor filter is proposed to suppress the background of seawater, enhance the target area of ships and detect and extract the targets. Firstly, the image is multi-channel Gabor filtering to obtain multiple filtered enhanced output images. Secondly, three kinds of filtering image enhancement effect evaluation indexes are defined and the fuzzy evaluation matrix is established for the output image. Thirdly, the output images are calculated according to the fuzzy evaluation matrix The fuzzy comprehensive evaluation value of each channel is selected and the output image with the best filtering enhancement effect of each channel is selected as the significant image of the channel filtering output. Finally, the corresponding fusion weight is calculated through the fuzzy evaluation values of each significant image, Significant images were weighted superposition and fusion, the ship target fusion enhancement images and detection. The experimental results show that the proposed method can adaptively select the Gabor filter output image with good background suppression and enhancement of the target region of the ship for fusion, and the fused image can effectively enhance the significance of the ship target. Compared with the existing detection method based on multi-channel Gabor filter, the target detection algorithm proposed in this study can effectively reduce the false alarm rate and improve the detection accuracy.