被动毫米波金属目标图像识别并自动统计出目标个数是**自动化技术的一个重要环节。给出了毫米波探测目标的简要原理，针对目前有关毫米波被动探测并自动计算出金属目标个数方法的欠缺，首次尝试并给出了被动毫米波金属目标图像处理方案及详细处理过程，在毫米波图像分辨率低的条件下，将图像二值化，用数学形态学对图像进行处理，用Canny算子进行边缘检测并提取图像的边缘轮廓，再用Visual C++6.0开发工具，实现二值图像连通域像素标记算法，搭建起金属目标个数的计数系统。给出系统的定量分析，并与面积法计算目标个数做了对比。经分析，系统具有准确性，且响应时间短。

To recognize metal targets image of passive millimeter wave (PMMW) and count the number of targets automatically is a key part of military automation technology.A brief principle of detecting target with PMMW was given.A scheme for PMMW image processing of metal targets and the detailed process were given.At presentthe PMMW imaging was in low resolution.In this schemethe image was transformed into a binary image.A method combining mathematical morphology was proposed to process the image.The Canny operator was used for detecting the edge of image and extracting image edge.A counting system was developed by Visual C++6.0 to count the number of metal targets.This system was based on connected area of binary image pixel labeled algorithm.The systems quantitative analysis was given and was compared with area method.Experiments show that this system is accurate and has short response time.