成像制导中的红外图像分割算法(英文)

Infrared image segmentation algorithm in imaging guidance

采用凝视焦平面阵列(FPA)的红外成像制导技术以其优越的性能在空空导弹中得到了广泛的应用。由于空中飞机目标表面各部位的温度不同,使得红外成像探测装置获得的目标图像的不同部位存在着较大的灰度差异,这给正确分割出目标带来了很大的困难。针对飞机目标红外图像的特点和成像制导技术中图像分割的目的,文中提出了一种有利于目标特征点识别的图像分割算法,它首先对目标红外图像进行阈值分割,然后选取合适的结构元素并利用数学形态学的并行性特点对目标内部的空洞区域进行填充,最后得到比较精确的分割图像,这种分割结果对基于灰度连续性的目标特征点识别算法非常有利。

IR imaging guidance technology that adopts Focal Plane Array (FPA) has been widely used in airtoair missiles by reason of its predominant capability. On the surface of aircraft, different parts are at different temperature, so the different parts of target image are different in grayscale. It will cause difficulty in accurate target segmention. Based on the features of aircraft IR image and the purpose of IR image segmentation in imaging guidance technology for new airtoair missile, a novel segmentation algorithm that is beneficial to target features recognition is proposed. The target image is segmented firstly. Then the areas which are inside of the target image are filled by selecting the suitable structural elements and utilizing the parallel characteristics of mathematical morphology. Finally, the accurate result of segmentation is obtained. The segmentation result of this method is of great help to grayscale continuity based on feature recognition algorithm.