CAJ | 학술논문

针对兼顾统计性和结构性纹理图像的分割存在的问题，提出了基于薛定谔方程的纹理图像分析与分割方法，实现了纹理图像有效分割，尤其是结构纹理图像的准确分割。该方法将纹理图像视作二维波函数，利用薛定谔方程获取纹理图像的势函数图像，然后采用耦合马尔科夫随机场与模糊聚类的算法分割函数图像。实验结果表明：文中给出的算法能够实现纹理图像较好的分割，特别是对于结构性纹理图像能获取较准确的分割。
침대겸고통계성화결구성문리도상적분할존재적문제，제출료기우설정악방정적문리도상분석여분할방법，실현료문리도상유효분할，우기시결구문리도상적준학분할。해방법장문리도상시작이유파함수，이용설정악방정획취문리도상적세함수도상，연후채용우합마이과부수궤장여모호취류적산법분할함수도상。실험결과표명：문중급출적산법능구실현문리도상교호적분할，특별시대우결구성문리도상능획취교준학적분할。
How to take into account both statistical and structural texture segmentation is an urgent problem.To resolve these problems, texture image analysis and segmentation method Based on schr? dinger equation were proposed.Its purpose is to realize the texture image segmentation effectively, especially structure texture image segmentation accurately. The texture image is regarded as a two-dimensional wave function in this method, and using schr? dinger equation for the potential function image, then using texture image segmentation algorithm based on fuzzy c-means and the coupled hidden markov random field.The experimental results show that the algorithm can get better results of texture image segmentation, especially structural texture image segmentation.