结合全局信息的LIF模型图像分割方法

A novel LIF level set image segmentation method with global information

针对LIF模型对初始轮廓敏感,CV模型对初始轮廓具有较强的鲁棒性,并且2种模型对噪声污染的图像不能取得令人满意结果的问题,在原先模型能量函数基础上,构造新的能量拟合项,增强对噪声的抗噪性.采用新的CV模型,使用图像全局信息得到粗分割结果.以粗分割轮廓作为新的LIF模型的零水平集,利用图像局部信息得到精确分割结果.同时使用一种新的边缘检测算子,重新定义边缘停止函数,进一步提高了模型的抗噪性.实验结果表明,它比CV模型、LIF模型、Chen模型和Qi模型更具优势,具有更强的抗噪性.

LIF model is very sensitive to the initial location outline,CV model is robust to the initial profile.two model cannot obtain satisfactory segmentation results for noise to solve these problems,a new energy fitting item is defined respectively based on the original model to enhance noise immunity.The improved CV model is employed to obtain coarse segmentation,the improved LBF model is employed to obtain accurately segmentation results with the initial contour based on the coarse result.A new edge detection operator is proposed to redefine the edge stop function to improve the noise immunity of the model.compared to the CV model,LIF model,Chen model and Qi model,the proposed model can get better segmentation results,with strong noise immunity.