在图像分类与识别算法研究中，目标几何特征的提取通常需要计算目标图像的最小外接矩形以获取长、宽等属性。针对该特点，

提出一种利用顶点链码与离散格林理论相结合的方式提取目标图像的最小外接矩形的算法。该算法只需根据顶点链中垂直或水平方向上的

点坐标即可求出目标的面积、形心和主轴。基于顶点链码和离散格林的主轴法和旋转法可快速求出目标的最小外接矩形。实验结果表明，

旋转法的运算速度是现有算法的 2 倍左右，主轴法的速度又比旋转法快速 2 倍左右-In the image classification and recognition algorithm, the extraction target calculate the geometric characteristics of the target image usually requires minimum bounding rectangle for the length, width and other attributes. For this feature, proposes a minimum bounding rectangle extraction target image algorithm uses vertex chain code and discrete Green' s theory combination. The algorithm can be calculated only according to the target chain, the vertex coordinates of the vertical or horizontal direction on the area and centroid of the spindle. Vertex chain code based and discrete Green' s spindle and spin method can quickly find the target minimum bounding rectangle. Experimental results show that the method of computing the speed of rotation of about 2 times the existing algorithms, surpasses the rotation speed of the spindle Rapid Method 2 times