基于二维最大熵和教与学优化算法的图像分割

摘要

As a global thresholding method,the two-dimensional maximum entropy method is used to consider the image gray level and spatial information,and the image segmentation results can be obtained in the case of low SNR.In order to improve the segmentation speed and efficiency in computation,on the basic of two-dimensional maximum entropy theory,the paper proposed a nonlinear inertia weight teach-learn-based optimization method to optimize the two-dimensional maximum entropy,the method take the 2-D maximum entropy asadaptive degree function of teaching-learning-based optimization algorithm,use the optimal threshold after optimizing to segment image.Due tothe nonlinear inertia weight teaching-learning-based optimization method has less parameter,and the convergence rate is fast,and the optimal segmentation threshold can be determined quickly through continuous optimization.The experimental results show that this method is not only fast and accurate,but also has strong adaptability.%作为一种全局阈值方法,二维最大熵综合考虑了图像灰度信息和空间信息,在图像信噪比低的情况下,也能得到理想的图像分割结果.为了提高图像分割的运算速度和效率,在基本的二维最大熵理论的基础上,提出使用非线性惯性权重的教与学优化方法对二维最大熵进行优化,将二维最大熵作为教与学优化算法的适应度函数,利用优化后的最优闽值对图像进行分割.由于非线性惯性权重的教与学优化方法参数较少,收敛速度快,通过连续优化,能较快确定最佳分割阈值.实验结果证明,所提图像分割方法不仅速度快、准确,而且具有较强的适应性.