基于区域活动轮廓模型的高光谱图像分割方法

Hyperspectral Image Segmentation Method Based on Region Active Contour

针对高光谱图像特点,提出了一种基于区域活动轮廓模型的高光谱图像分割方法。综合考虑高光谱图像的空间信息和光谱信息,对Chan-Vese方法中的能量函数加以改进,利用空间全局信息和同质区域的灰度一致性,约束能量函数空间项;利用目标光谱信息相似性,约束能量函数光谱项,最后通过能量函数最小化实现图像分割。该方法能够有效提取高光谱图像中的模糊轮廓,从而降低混合像元和目标周围阴影对分割造成的影响。利用两幅AVIRIS图像进行仿真实验,实验结果表明,提出的方法能够获得令人满意的分割效果,并且对复杂场景具有一定适应性。

Considering the characteristic of hyperspectral image, we have proposed a segmentation method based on region active contour. The energy function in Chan-Vese method is improved and both spatial and spectral information are employed. Spatial term of the function is restricted by global spatial information and the consistent intensity in homogeneous region; while spectral term is restricted by the consistent spectrum of target. Finally the image is segmented by minimizing the energy function. This algorithm can extract indistinct contours of hyper- spectral image, and thus reduce the influence caused by mixed pixels and the shadows around the target. Numerical experiments are conducted on AVIRIS data. Results show that this method reaches satisfying effects in performance and adapts to complex scene in some degree.