论文部分内容阅读
图像分割是对图像进行感兴趣区域提取与识别的基础,是图像分析的关键步骤。基于区域分割的JSEG算法是一种既融合了颜色信息又融合了空间信息的图像分割方法,在普通图像和视频图像中都能得到良好的分割结果。将这一算法引入到遥感影像的分割中,并对其做出改进,使之适用于多光谱遥感影像和纹理特征复杂的遥感影像的区域分割。实验结果表明,该算法很好的解决了由于影像中复杂的地物信息而产生的不同区域边界模糊的问题。
Image segmentation is the basis for image extraction and recognition of the region of interest and is a key step in image analysis. JSEG algorithm based on region segmentation is a kind of image segmentation method that combines both color information and spatial information. It can get good segmentation results in both normal and video images. This algorithm is introduced into the segmentation of remote sensing images and its improvement is made to make it suitable for the segmentation of remote sensing images with multi-spectral remote sensing images and complex texture features. Experimental results show that the proposed algorithm solves the problem of ambiguous boundaries between different regions due to the complex features in the image.