基于小波变换的图像过渡区提取与分割

Wavelet Transform based Image Transition Region Extraction and Seg mentation

复杂背景下目标的准确分割是自动目标识别中的关键技术之一。该文将基于过渡区提取的图像分割技术应用于自然纹理场景的图像分割,提出了基于小波变换的过渡区提取与分割算法,克服了以往过渡区直接提取方法的不足。实验结果表明,该方法用于自然纹理场景下的目标分割具有很好的分割效果。

Accurate segmentation of targets on complex background is one of the key techniques in Automatic Target Recognition(ATR).Transition region extraction based image segmentation technique is used to deal with segmentation problems on natural textual images in this paper.A novel wavelet transform based image transition region extraction and segmentation algorithm is proposed,which conquer the faults of the traditional transition region extraction method.Experi-mental results demonstrate that this method lead good result on target segmentation on natural textual background.