【摘要】 自动化苗木的嫁接，需要对苗木进行挑选。为此，设计一种苗木检测系统，通过传送带单株传送苗木，利用工业相机拍摄苗木的图像，并通过专门的开发识别软件，利用图像识别技术中依赖的OpenCV和用于数值计算的NumPy这两个外部库，对图片中的苗木轮廓提取，进行二值图像处理，利用格林公式计算出最小面积和长度及苗木的直径和弯曲度。测试结果表明：该方法有效，精度较高，可分析苗木的直径及弯曲度，剔除那些不适合自动化嫁接的苗木。更多还原

【Abstract】 The automatic seedling grafting, need to selection of nursery stock, design a kind of seedling test system, through the conveyor belt transmission per seedling, and using industrial camera images of the nursery stock, through specialized recognition software development, relying on the technology of image recognition of OpenCV and used for numerical calculation of NumPy both external libraries, seedling contour extraction of image, binary image processing, using the green’s formula to calculate the minimum size and length to calculate the diameter of the seedling and bending.After test this method is effective and high precision.Analyzed the diameter of the seedling and bending, excluding those grafted seedlings are not suitable for automation.更多还原