基于图像处理的菌落自动计数系统

Automatic colony counting system based on image processing

目的针对现有菌落计数仪结构复杂、便携性差等缺点,本文设计了一种基于图像处理技术和Android平台来完成菌落自动计数的系统。方法该系统由硬件拍照设备采集菌落图像,以智能手机为主要操作载体,对图像进行光谱阈值分割、中值滤波、洪水填充、开值运算、八邻域边界跟踪等多种算法处理后实现自动计数。结果与菌落计数仪相比,该系统使用移动端手机APP实时获取菌落图像并得到计数结果,可直接在手机上对结果进行手动校正,具有成本低、操作简单、便携性好等优点。结论基于图像处理的菌落自动计数系统与人工计数结果的相对误差不超过5%,满足菌落计数的要求。

Objective A system based on image processing technology and Android platform for auto counting colonies is designed,which overcomes the shortages of complicated structure and poor portability of colony counters. Methods The system contains a hardware photographing device to take pictures for colony samples,a mobile application to operate the system. The system uses image processing techniques of spectral threshold segmentation,median filtering,flood filling,open value operation combined with 8-connected boundary tracking algorithm for automatic counting. Results The system can show the colony pictures,count colonies in a mobile phone automatically,and correct the result manually if needed. The system has the advantages of low cost,simple operation,good portability compared to colony counters. Conclusions The relative errors of automatic counting compared with manual counting are less than 5%,meeting the requirements of colony counting.