基于红外图像的海上船舶分割算法研究

来源 :水雷战与舰船防护 | 被引量 : 4次 | 上传用户:along0429
下载到本地 , 更方便阅读
声明 : 本文档内容版权归属内容提供方 , 如果您对本文有版权争议 , 可与客服联系进行内容授权或下架
论文部分内容阅读
根据海面红外图像的特点,提出了一种基于红外图像的海上船舶分割算法。采用均值移动算法和霍夫变换检测海天线,以海天线为基准确定船舶出现的区域,采用区域分割算法分割船舶目标,根据区域生长法得到不同的目标区域,最后根据实验验证的本算法的有效性。
其他文献
The Tibetan plateau as one of the youngest orogen on the Earth was considered as the result of continent-continent collision between the Eurasian and Indian plates.The thickness and structure of the c
期刊
Based on the measurement of the arrival time of maxima magnitude from band-pass filtering signals which were determined using a new Morlet wavelet multiple-filter method,we develop a method for measur
期刊
Global Positioning System(GPS) and Interferometric Synthetic Aperture Radar(InSAR),used for monitoring crust deformation,are found to be very promising in earthquake prediction subject to stress-forec
期刊
According to the representation theorem of seismic energy radiation,we know that,at any point on the fault,the instantaneous seismic radiation energy,ES,the seismic moment M0,and the apparent stress σ
期刊
Noise correlation function(NCF) was calculated using the data of the Beijing Capital-Area Telemetered Digital Seismograph Network from June 12 to September 12,2005.Signal-to-noise ratio(SNR) is used t
期刊
一个月前的2月11日凌晨,郴州市安仁以及衡阳市部分地区的人们,纷纷传言当地突发"地震",结果引发恐慌,据媒体报道,有数万人当晚露宿街头。事后经当地政府及时辟谣,证实当时出现的一些房屋摇晃等特异现象,不过是衡阳某驻军的军事演习所致。
期刊
The 3-D crustal structure of P-wave velocity in East China is studied based on the data obtained by wide-angle seismic reflection and refraction surveys.The results suggest that a deep Moho disconti-n
期刊
对互花米草进行了40cm水深的持续淹水胁迫,研究了其生理指标的响应.结果显示,淹水胁迫下,互花米草叶片中可溶性糖含量和叶绿素含量在初期均显著低于对照,随后逐渐积累并高于对照.丙二醛含量在初期显著高于对照,随后与对照组的差异逐渐减小,至生长季末与对照水平相当.过氧化物酶活性在淹水后63d内均高于对照,至生长季末则低于对照,而超氧化物歧化酶活性则始终高于对照.至生长季末,持续淹水胁迫条件下的互花米草生
期刊
Soil moisture distribution shows highly variation both spatially and temporally. This study assesses the spatial heterogeneity of soil moisture on a hill-slope scale in the Loess Plateau in West China
期刊
提出了一种新颖的正则化方法—判别性正则化(Discriminative regularization,DR),为分类提供了一种通用的结合样本先验信息的方式。通过将先验信息引入到正则化项中,DR不但使分类器实际输出与期望输出之间的经验损失达到最小,而且能在输出空间中同时最大化类间散性与最小化类内紧性。此外,通过将等式约束嵌入到目标函数中,DR的求解还可转化为解线性方程组问题,从而得到全局解析解。分类
期刊