【摘要】 针对传统目标检测方法只能应用在缓慢变化背景中的不足,文章从分析背景与目标的不同成像特性入手,提出了一种复杂海面背景下的红外舰船目标检测方法.该方法利用成像目标与背景的不同梯度特性,构造多向梯度检测函数来区分目标与背景,并对多向梯度算法执行结果利用灰度形态学滤波来进行背景泄漏噪声抑制.利用实际拍摄的红外图像对该算法进行了实验并与传统分法进行了比较,结果表明利用该方法能有效地检测出高起伏海面背景下的舰船目标.更多还原

【Abstract】 The traditional objects detection algorithms can only work on stationary Background.To overcome this problem,this paper gives a detection method for infrared chip target in a complex sea environment based on analysis the difference of target and background.This method constructs a multi-orientation gradient detection function to detect target from background and uses grayscale morphology method to wide off the background noise.This method is evaluated on real infrared image.Experimental results show that this method has high performance compared with traditional moving objects detection algorithm and can detect the chip target easily.更多还原