快速高维均值平移算法及其在航空图像分割中的应用

随着军事伪装技术的广泛应用,如何有效提取与背景色彩接近的目标是提高航空图像目标识别能力的关键。针对背景色彩接近目标的航空图像分割问题,应用一种基于快速高维均值平移的图像分割方法对航拍图像进行处理。在包含有空间,色彩和纹理的高维特征空间中对图像进行均值平移滤波,再通过聚类方法分割图像。同时,针对特征空间维数增大带来的大计算量问题,对均值平移算法进行了改进,给出了一种基于均值平移向量预测的快速算法。实验结果表明,算法用于航拍图像分割,分割效果和运算速度均优于标准算法。

Application of Fast High-dimension Mean Shift Method to Aerial Image Segmentation

Extraction of target with similar color to the background becomes increasingly important in aerial target recognition with the development of military camouflage technology.For this problem,an improved image segmentation algorithm based on fast high dimension mean shift is studied here.Firstly,the proposed method was used for image filtering in a high dimension space with spatial,range,and texture features.Then,the image is segmented by clustering algorithm.Furthermore,fast mean shift method based on predic...