基于局部和全局信息的自适应水平集图像分割

发布时间：2018-10-16 16:56

【摘要】：针对仅采用局部或全局信息无法快速准确分割灰度不均匀图像的问题,提出了一种基于局部和全局信息的自适应水平集图像分割模型。首先,利用图像局部信息和全局信息建立局部能量项和全局能量项,并且利用演化曲线轮廓内外小邻域的灰度均值差作为自变量,建立了权重函数模型,实现了局部能量项和全局能量项之间权重的自适应调整,提高了模型分割灰度不均匀图像的效率和准确性。其次,提出了一种新的能量惩罚项,避免了水平集函数的重新初始化,增强了数值计算的稳定性。最后,为验证模型的优越性,将模型与CV模型、LBF模型和LGIF模型进行了对比,并通过分割时间、迭代次数以及相似度等指标对分割结果进行了客观、定量分析。最终结果表明:该模型不但对初始轮廓具有较高鲁棒性,而且对灰度不均匀图像具有较高的分割准确性与分割效率。
[Abstract]:An adaptive level set image segmentation model based on local and global information is proposed to solve the problem that only local or global information can not be used to segment gray-scale heterogeneous images quickly and accurately. Firstly, the local energy term and global energy item are established by using the local and global information of the image, and the weight function model is established by using the gray mean difference of the small neighborhood inside and outside the evolution curve contour as the independent variable. The adaptive adjustment of weights between local and global energy items is realized, and the efficiency and accuracy of model segmentation are improved. Secondly, a new energy penalty term is proposed, which avoids the reinitialization of the level set function and enhances the stability of the numerical calculation. Finally, in order to verify the superiority of the model, the model is compared with the CV model, the LBF model and the LGIF model, and the segmentation results are analyzed objectively and quantitatively by the segmentation time, iteration times and similarity. The results show that the proposed model is not only robust to the initial contour, but also has high segmentation accuracy and efficiency for gray-scale non-uniform images.
【作者单位】：西北工业大学自动化学院;西安交通大学人工智能与机器人研究所;
【基金】：国家自然科学基金项目(61473219)
【分类号】：TP391.41

【参考文献】