中文摘要：

为减小基于机器视觉的农产品分选机的体积、降低设备成本,提出了一种以DSP为视觉核心、ARM为控制核心的小型农产品分选机。为此,详细介绍了基于机器视觉技术的小型农产品分级机控制系统及控制方法,并以圣女果、金桔及桂圆干等3种物料作为试验对象,测试系统的可靠性和工作效率。结果表明：针对家庭农场及小规模农产品的处理和营销,在基于机器视觉的小型农产品分选机中采用ARM与DSP相结合的方法进行分选是可行的,能够替代PC机或工业计算机实现分选机的低成本和小型化。

英文摘要：

In the aims of lowering the size and cost of agricultural products sorting machine, we raised an small-scale sorting machine with DSP-based vision kernel and ARM-based controlling kernel. The control system and the control method of the machine vision-based agricultural product sorting machine were introduced in this paper, and cherry toma- tos, cumquats and dried longans were used to estimate the reliability and working efficiency of the system. The results showed feasibilities of machine vision-based agricultural product sorting machine featured with ARM and DSP in the aim of the processing and merchandising of small-size agricultural products, together with replacing PC and industrial com- puters in order to achieve lower size and cost of sorting machine.