基于MATLAB/GUI的无人机遥感图像分类系统设计

为了提高无人机遥感图像分类技术在复杂地形中的勘察效率，设计了一种基于MATLAB/GUI的无人机遥感图像分类系统，实现了图像分类的可视化操作。采用K-Means聚类与Otsu阈值分割相结合的方法对无人机遥感图像进行分割与分类处理，并对初分割结果进行边缘检测、噪声滤波和形态学优化等图像处理，精确提取各类地物边界，提高图像分类精度；对分类处理后的结果进行矢量化输出，提高了本系统与其他软件的交互性；根据航拍高度、镜头焦距和传感器尺寸等信息，可计算每一类地物的实际面积，实现区域面积大小的快速统计，促进了无人机遥感技术在公路地质选线中的应用。

Design of UAV Remote Sensing Image Classification System Based on MATLAB / GUI

To improve the survey efficiency of UAV remote sensing image classification technology in complex terrain， a MATLAB/GUI-based UAV remote sensing image classification system was designed to realize the visual operation of image classification. Based on the K-Means clustering and Otsu threshold methods， UAV remote sensing images were segmented and classified. After that， to extract boundaries of various ground features and improve the accuracy of image classification， image processing algorithms such as edge detection， noise filtering and morphological optimization were conducted on the initial segmentation results. Finally， the classification results were vectorized and output， which improves the interactivity between the designed system and other software. According to information such as aerial height， lens focal length and sensor size， the actual area of various types of surface objects can be calculated， and statistics of area size can be achieved efficiently， which promotes the application of UAV remote sensing technology in highway geological route selection.