马铃薯质量和形状自动分级系统研究—基于计算机视觉

Research on Automatic Grading System of Potato Quality and Shape—Based on Computer Vision

马铃薯是一种粮食和蔬菜作物,具有很高的产量和经济价值。我国的马铃薯种植面积和产量均居世界首位,但是检测分级主要由人工完成,作业效率较低,准确性和客观性差。随着科学的发展和技术的进步,我国开发出了多种新的马铃薯检测分级方法、系统和装置,以计算机视觉技术应用最多。为此,基于计算机视觉,设计了针对马铃薯质量和形状的自动分级系统,可以实现40个/min的自动分级速度。系统对单薯的质量和形状分级效果均较好,总体准确率分别达到95%和90%,可以作为提升我国马铃薯产业水平的技术支撑。

Potato is a food crop and vegetable crop with high yield and economic value. China＇s potato planting area and output rank first in the world, but the detection grading mainly by the manual completion of the work efficiency is low, accuracy and objectivity is poor. With the development of science and technological progress, China has developed a va- riety of new potato detection grading methods, systems and devices, of which the most used computer vision technology. Based on computer vision, this paper designs an automatic grading system for potato quality and shape, which can achieve automatic grading speed of 40 / min. The quality and shape grading effect of the whole potato is good, and the overallaccuracy rate is 95% and 90% respectively. It can be used as a technical support to enhance the level of potato industry in China.