声呐图像背景灰度分布特性分析及模型研究

Background gray level distribution in sonar images

在图像声呐滤波、分割以及目标识别中,像素灰度分布特性及其模型的选择与建立具有非常重要的作用。针对当前声呐图像背景灰度分布模型研究尚不够深入的现状,分别采用威布尔分布和瑞利分布模型对声呐图像背景区域像素点灰度分布进行了仿真验证。结果表明,威布尔分布较瑞利分布来说能够更好地逼近图像声呐灰度分布曲线,且具有较强的自适应性。但在某些情况下,可采用瑞利分布替代威布尔分布,以提高运算效率。

Distribution of gray levels is important in filtering, segmentation and identification of sonar images. This paper uses the Weibull law and Rayleigh law for sonar imagery to study the problem. The results indicate that the Weibull law is more appropriate than the Rayleigh law since the former has stronger adaptability. Under some special conditions, however, the Rayleigh law may be used instead to improve the processing efficient.