基于二维Otsu的舰船目标图像分割方法  

Image Segmentation Based on Two-Dimensional Otsu Algorithm

作　　者：余扬 YU Yang(Combat Service Support Company of Navy Staff in Southern Theater,Zhanjiang 524000)

机构地区：[1]南部战区海军参谋部作战勤务保障大队,湛江524000

出　　处：《舰船电子工程》2022年第1期36-39,共4页Ship Electronic Engineering

摘　　要：针对图像分割问题,论文研究提出了一种基于局部网络区域划分技术的二维Otsu阈值分割方法。首先,该方法采用整体到局部的研究思路,基于原始图像的二维直方图划分为多个局部矩形子区域,并将子区域视为一个独立的像素点,构建一个新的二维直方图,实现阈值的初步计算;其次,由阈值的计算结果,确定其所在的区域,以此建立阈值求解模型;最后采用基于粒子群多目标优化模型对阈值进行二次计算,即获取最终结果。理论分析和试验结果表明,该算法的时间复杂度降低为O(L)(L表示图像的像素灰度级),且获得了较好的分割效果。For image segmentation,a fast two-dimensional Otsu threshold segmentation method based on local grid region di⁃vision technique is proposed.In this method,firstly,by utilizing coarse-to-fine region division idea,2D histogram of original im⁃age is divided into regions,and each region is used as a point to form a new 2D histogram,to which the threshold is computed for the first time.Then according to the first threshold computed result,the region where the threshold is located is determined,and in this region,the second threshold solution model is built.Finally,particle swarm optimization(PSO)algorithm is applied to get the final threshold.Theoretical analysis and experiments show the method time complexity is reduced to O(L)(L is the pixel gray lev⁃el),and simultaneously obtain better segmentation results.

关 键 词：图像分割 局部盒滤波 OTSU算法 二维直方图 

分 类 号：TP751[自动化与计算机技术—检测技术与自动化装置]