红外技术, 2018, 40 (6): 569, 网络出版: 2018-08-04
基于快速杂交拟合能量模型的红外海面目标检测
Infrared Sea Surface Target Detection Based on Fast Hybrid Fitting Energy Model
杂交拟合能量模型 水平集方法 红外成像 海面目标 检测概率 hybrid fitted energy model level set method infrared imaging sea surface target detection probability
摘要
针对传统的红外海面目标检测方法所存在的缺陷,即很难准确的捕获模糊目标区的外部轮廓,提出了一种基于快速杂交拟合能量模型的红外海面目标检测方法。该方法有效的融合了小波多尺度分解、海天线提取、多级滤波及杂交拟合能量模型。在将图像数据牵引至杂交拟合能量模型所对应的演化过程之前,经历了局部子带内的多级多带宽的滤波处理,最后将子图内的检测结果映射到原始的图像数据上。定量地评价了所提算法在信杂比增益、背景抑制因子及检测概率等指标上的性能。实验结果表明,相比传统的目标检测算法,该方法获得了更高的检测概率。
Abstract
Aiming at a problem of traditional infrared sea surface target detection methods, to accurately capture the outer contour of the blurry target region, a detection method based on fast hybrid fitting energy model for sea surface targets is proposed. The method combines the wavelet multiscale decomposition, the sea-sky-line extraction, the multilevel filter, and the hybrid fitting energy model. Before carrying the image data to the evolution process corresponding to the hybrid fitting energy model, multi-level multi-bandwidth filtering is processed in a local band. Finally, the results detected in the subimage are remapped to the original image data. The test results of real infrared image sequences demonstrate the validity of the proposed method.
赵楠楠. 基于快速杂交拟合能量模型的红外海面目标检测[J]. 红外技术, 2018, 40(6): 569. ZHAO Nannan. Infrared Sea Surface Target Detection Based on Fast Hybrid Fitting Energy Model[J]. Infrared Technology, 2018, 40(6): 569.