基于阴影搜索法的飞机目标遥感图像分割研究

Study on Segmentation of Airplane Object in Remote Sensing Based on Shadow-Searching Method

介绍阈值分割的原理并分析传统图像分割中阈值的确定方法;为提高对飞机目标遥感图像的分割效果,采用阴影搜索法对图像阈值进行判定。该算法对图像进行横向或纵向扫描,根据像素的变化频率判定阴影区域并确定目标与背景的关系,实现图像分割中最优阈值的自动提取。该算法克服了传统方法灵敏度低对分割效果的影响,有很好的通用性和有效性。

Principle of threshold segmentation is presented and confimling threshold methods in traditional image segmentation are analyzed. In order to enhancing effect, the shadow - searching method is put forward in deciding threshold of image. The method scans the image from landscape orientation and portrait ,decides shadow area based on changing frequency of pixel and confirms relation of object and background, thereby, achieves automatic extraction of furthest threshold in image segmentation. Threshold segmentation is done on the base of shadow- searching method in remotely sensing imagery. Experiments show that new arithmetic conquers delicacy degree lowness of tradition methods that impact segmentation. With other threshold segmentation arithmetic comparing, new arithmetic has better currency and validity.