摘要
复杂背景下的红外弱小目标检测在红外图像处理领域中具有重要价值。本文提出了一种以背景抑制与目标增强相结合,使用国产多核DSP并行运算的红外弱小目标检测方法。通过高斯滤波和梯度锐化有效的提升了信噪比,理论分析及实验结果表明,本算法能够提高检测率、降低虚警率、具有很强的检测鲁棒性和检测实时性。
Infrared dim targets detection under the conditions of complicated background is useful in infrared image processing fields.The article offers an method of small targets detection in infrared images with background suppression and targets enhancement,which deals under the nationalization of multi-core DSP.The method improved the ratio of signal-noise with Gaussian filter and gradient sharpening,the theoretical analysis and experiment results indicate that the proposed method has good performance of increasing detection rate and decreasing false rate,and has good robust and detection in real time.
作者
孙秀峰
SUN Xiu-feng(Sichuan Changhong Electronic Technology Co.,Ltd.,Mianyang Sichuan 621000)
出处
《数字技术与应用》
2020年第10期80-82,共3页
Digital Technology & Application
关键词
国产化
红外弱小目标
高斯滤波
nationalize
infrared dim targets
Gaussian Filter