基于改进CV模型的图像分割算法

Image segmentation algorithm based on improved CV model

传统CV模型在目标图像存在噪声干扰及图像背景较为复杂的情况下,图像分割效果较差,极易造成误分割。为了提高基于CV模型图像分割的分割效果及分割效率,提出一种基于改进CV模型的图像分割算法。首先,根据曲线演化理论对CV模型的曲线驱动力进行简化,以此提高模型的分割效率;然后,利用L1范数构造CV模型的能量泛函,同时引入中值替代传统CV模型中的曲线拟合中心,在简化数据计算的同时,提高模型对噪声的鲁棒性;最后,将该曲线驱动力与L1范数能量泛函进行融合,以此构造最终的改进CV模型的能量泛函。将所提模型与传统CV模型、LIF模型、局部二值模型以及偏置场修正水平集模型的实验结果进行对比,结果表明所提模型分割效果最优,且分割速率最高。

The traditional CV(Chan-Vese)model has poor image segmentation effect and may result in segmentation mistake in the cases of noise interference and complex image background existing in the target image.An image segmentation algorithm based on improved CV model is proposed to improve the segmentation effect and segmentation efficiency of the model.The curve driving force of CV model is simplified according to the curve evolution theory,so as to improve the segmentation efficiency of the model.The L1 norm is used to construct the energy functional of CV model,and the median value is introduced into it to replace the curve fitting center of traditional CV model,which can simplify the data calculation,and improve the model robustness against noise.The L1 norm energy functional is fused with the curve driving force to construct the energy functional of the improved CV model.In comparison with the experimental results of the traditional CV model,LIF model,local binary model and bias field correction level set model,the proposed model has the best segmentation effect and fastest segmentation speed.