基于图像处理技术的西洋参外形特征参数测定

Study on the Measurement of Shape Parameters of American Ginseng Based on Image Processing Technology

采用数码相机在特定环境下拍摄西洋参样本图像,在Matlab2010软件平台下编写相应的程序,首先对原始图像进行滤波、目标与背景分割、二值化及边界检测等预处理,接着通过参照物确立像素点与物体实际外形参数的比例关系,最终建立西洋参长度与横切面直径的测定模型.实验结果表明,所建立的测试模型具有较高的测量精度,达93.8%,与人工测量相对误差在6.2%范围之内.

Using digital camera in the specific environment of ginseng sample image to image,under the Mat-lab platform,a variety of image processing methods based on the image in object extraction,binarization andboundary detection etc pretreatment,through the establishment of reference points and the actual objects likeshape parameters of proportional value,finally established American ginseng length and transverse section di-ameter measurement model. Experiment results show that accuracy of 93.8%,and artificial measuring relativeerror in the range of 6.2%.