为实现合成孔径雷达（SAR）图像分类算法的测试与评估，将VC/Matlab 混合编程技术应用到SAR 图像分类平台开发中，分析与

比较了4 种VC/Matlab 混合编程方法及各自优缺点，并着重研究了基于组件对象模型（COM）的VC/Matlab 混合编程方法。使用Matlab

COM 编译器创建了SAR 图像分类算法组件，在VC 中调用其导出的接口函数。在此基础上，完成了SAR 图像分类平台的实例开发且

可脱离Matlab 环境运行。实验结果表明，该方法较好地发挥了VC 与Matlab 各自的优势，提高了SAR 图像分类平台开发的效率。-In order to test synthetic aperture radar （SAR） image classification algorithm，VC/Matlab mixed programming technology was

used in SAR image classification platform development. Four VC/Matlab mixed programming methods were introduced and analyzed.

Comparing with other three methods，the method based on component object model （COM） was studied particularly. SAR image

classification algorithm component was created by Matlab COM complier，and its interface function could be invocated with VC. Using this

method，a SAR image classification platform was realized. The platform could run without Matlab. The experimental result shows a good

combination of VC and Matlab. So，developing efficiency could be enhanced a lot.