【摘要】 研究了红外序列图像中弱小目标的检测。对图像序列进行小波变换,利用目标小波变换的特性,对单帧小波变换后的各高频分量进行累加,利用门限处理并通过与一个与管道得到包含候选目标点的图像,最后利用目标运动的连续性,进一步检测,得到目标点的位置。实验结果表明,该方法能有效地检测红外弱小目标。更多还原

【Abstract】 For the detection of dim moving targets in the infrared image sequences,the wavelet transform is presented here.According to the features of the dim moving targets after wavelet transform,an integration of the three parts of the high frequency domain is needed.After a threshold and an and-pipe,an image with the potential targets and noise is obtained.At last,by using the continuity and regularity of motion for target and the randomness for noise recognizes the targets in the images are recognized.The experiment result indicate that this method can detect and locate small moving target validly and is robust respect to noise interference.更多还原