聚类算法在航空图像颜色处理中的应用

Application of Cluster Algorithm in Aerial Images Processing

将聚类算法以及颜色传输算法应用到航空图片颜色处理中。首先,利用聚类算法对两幅不同环境下的航空图片进行聚类分析,产生若干聚类域;然后,利用颜色传输算法实现匹配颜色聚类域之间的对应传输;最后将传输结果转换到RGB空间显示。为了确定两幅图像间的传输方向,为每个聚类域设置一个匹配因子,并根据不同的图像特征采用不同的匹配因子计算方法;然后选择两幅图像中匹配因子最接近的域作为匹配域,实现对应传输。实验表明,该算法能够获得不同环境下的航空图片。

Clustering algorithm and color transferring were applied in aerial images processing.First,fuzzy color clustering was applied to obtain different clustering sets.Then,corresponding transferring between matching clustering fields was realized by color transferring algorithm.Finally,transferring results were turned to RGB place.To determine transferring direction,matching factor for each clustering field and computes the matching factors according to different target images are set.Then selecting a field in the source that is the most approximate to the current field in the target as the matching field.The experimental results show that the algorithm proposed in the paper can get desire aerial images in different environment conditions.