Abstract: To solve high misjudgment rate and inefficient for detection pesticide residues of a large quantity of vegetables, a fast and accuracy method based on machine vision is proposed to judge rapid test card result. Firstly, image acquisition system is designed to acquire rapid test card image. Then, the rapid card image is processed using edge detection technique and characteristics extraction. Experimental results show that the method can judge about 500 samples in one minute. The present method is to meet the requirements of large quantities of vegetable pesticide resresidues and is helpful for the construction and management of green vegetables traceability system.

Key words: machine vision, pesticide residues, rapid detection

摘要： 针对人工判读大批量农药残留速测卡结果存在的误判率高、效率低等问题，提出基于机器视觉的农药残留速测卡结果准确快速判读方法，根据判读原理设计图像采集系统，利用边缘检测技术结合特征提取对速测卡图像处理分析。实验结果表明，该方法在保证准确性的前提下，判读效率达到约500次判读每分钟，能够满足蔬菜采摘前和销售前大批量农药残留检测需求，有助于绿色蔬菜溯源系统的建设和管理。

关键词: 机器视觉, 农药残留, 快速判读