摘要
阈值方法是一种重要的图像分割方法,在图像分割中得到了广泛应用。Otsu算法虽然是图像分割阈值法中较好的方法之一,但是由于传统的Otsu算法通常用穷举法求解,使得处理多阈值问题时运算速度太慢,难以满足应用需求。为了快速有效地确定阈值,提出了一种改进的Otsu算法。将Otsu算法转化为一个非线性0-1数学规划问题,再利用遗传算法求解得到最优阈值。通过对测试图像的分割实验,表明该算法与传统的Ot-su算法相比运算速度有非常显著的提高,能够满足一般的应用需求。
The thresholding for image segmentation is an important and well-established method that has been widely applied to this problem. Conventional Otsu algorithm is however, computational suffering for using the exhaustive searching strategy to find the optimal thresholds. It is thus inapplicable in the selection of multilevel thresholds. In this paper, a modified Otsu method is proposed to determine the thresholds with improved efficiency. This is accomplished by transforming the Otsu method to a nonlinear 0-1 programming problem, which can be solved by genetic algorithms. The results on the testing images show that the computational speed of the proposed method is significantly improved to accommodate the general use of image segmentation.
出处
《计算机工程与应用》
CSCD
2012年第10期197-199,232,共4页
Computer Engineering and Applications
关键词
图像分割
多阈值
0-1规划
image segmentation
multilevel thresholding
0-1 programming
