摘要
DBC和MDBC估计分数维时信息丢失较多,导致估计精度降低和范围缩小。针对此局限性,提出了利用图像灰度均值反映整体特征的GDBC分数维算法,结合实际SAR舰船图像目标识别的应用,从分数维估计、拟和误差计算和动态范围测量等方面,将GDBC和其它计算方法进行了比较和分析。实验结果表明该算法更加有效,覆盖范围更广。
DBC and MDBC lose a lot of information when estimating the fractal dimension of an image. In order to overcome sue defect, a modified DBC, named GDBC is proposed, which takes use of the mean gray level of an image. Experiments are designe in three aspects to compare GDBC and the other two methods. The results suggest that GDBC is better than other two.
出处
《计算机工程与设计》
CSCD
北大核心
2005年第2期315-316,325,共3页
Computer Engineering and Design
基金
武器装备预研基金项目(403010503)。
