改进的机场停泊飞行器目标分割算法

考虑到飞行器目标在整幅图像中所占的比例往往较小,且图像背景复杂,本文提出了一种基于机场区域提取的飞行器目标分割算法.该算法首先利用Hough变换检测直线的特性,定位机场跑道和停机坪的位置,并结合教学形态学等图像处理技术去除了非机场区域;在提取机场区域后,再选择适当的阈值对图像进行分割,最后经过形态学去噪、小区域去除等步骤分割出飞行器目标.实验结果表明,该算法改进了以往机场区域提取算法保留候机楼等附属部分以及提取结果中存在机场区域以外区域的缺点,较好地实现了机场停泊飞行器目标的分割,为下一步准确识别飞行器类型奠定了基础.

Improved Target Segmentation Algorithm for Aircraft Parking at Airport

Considering that the aircraft targets have less scale on the whole image with a complicated background,an algorithm is proposed to segment aircraft targets based on the airport area extracting.Firstly,by using the features of the detecting lines of the Hough transform,the runway and the apron are located.Then,the area out of the airport are wiped off by using some image processing technologies,such as mathematical morphology.After extracting the airport area,a proper threshold value is selected to segment the image.Finally,the aircraft targets are segmented after some steps,such as morphological de-noising and small area removing.The experimental result shows that the algorithm can overcome the disadvantages of the previous algorithms of the airport extraction for retaining the subsidiary parts,such as terminal buildings,thus the extracting results contain area out of the airport area.