Research of image segmentation method based on enhanced course TLBO
To improve the quality of the segmented images,and the efficiency of computation,an image segmentation method based on enhanced teaching and learning based optimization(ETLBO)is proposed.The traditional TLBO algo-rithm is improved with the proposed reverse learning technique and small probability mutation strategy,and the global search speed and optimization accuracy are enhanced while retaining the advantages of low computation cost and high sta-bility.In addition,the concept of minimum cross entropy(MCE)is used to transform the thresholding problem of image segmentation into an optimization problem.The optimal combination of multiple thresholds is obtained with ETLBO,which can minimize the cross entropy between the original image and the segmented image,therefore improving the visu-al quality of the segmented image.The experimental results indicate that the proposed method outperforms other ad-vanced methods in terms of the uniformity and fitness values,and effectively reduces computation time.
NETL
NSTL
万方数据
