【作者】 郭海涛； 王连玉； 田坦； 戴愚志； 张春田；

【Author】 GUO Hai-tao1,2 , WANG Lian-yu2 , TIAN Tan3 , DAI Yu-zhi4 , ZHANG Chun-tian1(1. School of Electronic Information Engineering, Tianjin University, Tianjin 300072,China; 2. Laboratory of Underwater Acoustic Engineering, National Ocean Technology Center, Tianjin 300111,China; 3. College of Underwater Acoustic Engineering, Harbin Engineering University, Harbin 150001, China; 4. School of Civil Engineering, Tianjin University, Tianjin 300072,China)

【机构】 天津大学电子信息工程学院； 国家海洋技术中心水声技术研究室； 哈尔滨工程大学水声工程学院； 天津大学建筑工程学院；

【摘要】 给出了二维属性直方图的概念,在此基础上提出了一种基于二维属性直方图和遗传算法的图像自动阈值化方法.该方法对二维阈值进行编码,根据二维属性直方图的Otsu算法确定适应度函数,通过遗传计算确定最佳分割阈值.将该方法用于一种海底小目标图像阈值化,经过240次适应度函数的计算即可得到最佳分割阈值.结果表明,该方法适用于直方图不是理想双峰形状的图像,比基于二维属性直方图的Otsu算法速度更快.更多还原

【Abstract】 The concept of the two-dimensional bound histogram (TDBH) was given. An automatic threshold selection method based on the TDBH and genetic algorithm was presented. With this method, the two- dimensional thresholds would be coded, and the fitness function would be established according to the Otsu algorithm based on the TDBH. The optimized two-dimensional threshold for segmentation was determined by genetic computing. The proposed method was used in threshold determination for the image of a small underwater target. The optimized two-dimensional threshold would be obtained after the value of the fitness function was calculated 240 times. The results show that the method is well applicable to the image with a non-ideal bimodal histogram, and that it is faster than the method based on the TDBH.更多还原