Algorithm of small target detection based on one dimensional average gray absolute difference maximum

It is an unfathomed and difficult problem that weak and small targets are automatically detected in complicated background and low SNR(signal noise ratio).One dimensional average gray absolute difference maximum(1-D AGADM) was presented as the dissimilarity measurement between target and background region.The filter based on 1-D AGADM was used as preprocessing method for small targets detection.It not only enhanced small target,but also suppressed the clutter and noise.Then,the theory of background forecast was utilized to detect small targets.Experiments indicate that the presented filter greatly improved the SNR of small target image,and ensured the veracity of background forecast algorithm in the following detecting small target.Moreover,compared with the 2-D AGADM filter,the novel filter can adapt to the shape of target better and the processing speed is faster.