基于DSP和FPGA的图像选别系统设计  

Design on Image Sorting System Based on DSP and FPGA

作　　者：赵志衡[1] 宋欢[1] 马涌[1] 张懿夫 

机构地区：[1]哈尔滨工业大学电气工程及自动化学院,哈尔滨150001

出　　处：《计算机测量与控制》2017年第8期123-126,共4页Computer Measurement &Control

摘　　要：色选设备是集光、电、气、机及图像处理为一体的高科技综合技术;基于DSP和FPGA的图像选别系统具有精度高、处理速度快、集成度高、功耗小等优点;首先介绍图像选别系统的硬件组成,然后以破损花生检测为研究对象,通过MATLAB编程研究不同处理算法的优缺点,最后移植到DSP上测试编写的程序,提出破损花生的检测算法,实现了依据破损面积大小对花生的选别;在花生筛选中,除了破损花生检测,还有花生的大小检测,因此在上述基础上进一步提出花生几何参数的计算方法。Color sorting technology is a highly-synthesized one connected with light,electricity,gas,machine and image processing.Sorting systems based on DSP and FPGA have many advantages such as high precision,high processing rate,high integration level and low power consumption.It first introduces the hardware structure of the image sorting system,and then uses MATLAB simulation tool to research several images processing algorithms.Finally,it tests the programs of the sorting system on DSP and proposes the broken peanuts detection algorithm.In this way,the selection of peanuts according to the size of damaged area is achieved.Except the detection of damaged peanuts,the target of peanuts screening also includes the size of peanuts.Based on the above,it puts forward the calculation method about the geometric parameters of peanuts.

关 键 词：机器视觉 图像处理算法 图像选别系统 

分 类 号：TP393[自动化与计算机技术—计算机应用技术]